Home

ClearPath OS 2200 TeamQuest Baseline and TeamQuest Online

Contents

1. s system name where f specifies the names of the system log files to be processed systemlog is the name of the system log file n is the file cycle number When processing multiple system log files multiple f options are needed one for each cycle to be processed All files must be in sequential order S specifies a system name The system name parameter specifies the name of the system source of the system log files It determines where to store the data in the performance database The system name parameter is case sensitive TQ 01011 18 9 35 Customizing Data Collection Probes 9 36 The tqsys d and a parameters are also required for system log file input see 9 3 1 Only one aggregation set a can be specified when using the f parameter If you do not want tqsys to load the system log data into the version of the performance database specified on the Database setting in the tq ini file use the d parameter to specify an alternate database version Note Before loading data into an alternate version of the performance database you must use the tqdbu utility to create this alternate version see 10 4 If the b and e parameters are used with the f parameter tqsys processes the information in the specified log files for the date range specified on the b and e parameters The tqsys program processes each system log file storing data points in the performance database as specified by the aggregation set
2. Maintaining Derived Statistics with tqzip eeee Unloading Derived Statistic Definitions Loading Derived Statistic Definitions 0 0 Customizing Data Collection Probes INTROGUCTION sirian acini een einen ieee toctl Runstream vioista aatas di edna SIP Data Collection ccccccccccesccccecececceeesessssteaaeeeeeeeees tqbsp Processor Call Statement eee Terminating Data Collection eceeeeeeeereees Abnormal Termination of Data Collection _ SIP File Input arrra ete raet Sre ddaki irura darpa adian System Activity Monitor Collection ccccccesseeeeeeees tqsam Processor Call Statement Terminating Data Collection eceeeeeeeseeees Abnormal Termination of Data Collection _ Workload Characteristics for the System Activity Monitor Probe ou eceeeececesssseseeseeeceeeseseeeeeeeeeees tasam Table Data o cccceeeeeeceeeeeeeeeeeeeeeeeeeeeeeeees Runs Table Information ccccceesessseseeeeees Transactions Table Information 7 27 7 28 7 30 7 31 7 82 Peer ee PERT P LLLLLLL PPPLPPP ankwown WAODNAF o L NO 8 17 8 18 8 21 PEPEPEPE ee Pear NOOTRWANND PP PP o N TQ 01011 18 Contents TQ 01011 18 9 5 MIPS Metering Data Collection oo eccceeeeeeestteeeeeees 9 5 1 tqmips Processor Call Statement ceeeee 9 5 2 Terminating Data Collection ceeeeeeeeenee
3. TeamQuest View Software essssseeeeeeseeesesrrenes Product Compatibility 2 ceececeeeeceeeeeeeeeeeeeeeeeeeeeeeeees Section 2 TeamQuest Baseline and TeamQuest Online Concepts 2 1 2 2 2 2 1 2 2 2 2 2 3 2 2 4 TQ 01011 18 OVOIVIOW aa E E EE E E A EE Performance Database auu ccceceseseseeceeeeceeeeeeeeeeeeesaeaes Aggregation Sets oo eececeeececesseeeeeeeeeeseteeeteneeeeees Parameters e N a e a Data Values Sine A a denotes Table Data iaat aaraa ieaS Data Collection Srna n a a bias dee Workload Management sssssssssssisssesrrrssrerrrsssrrrnnsseens Alarms and Thresholds cccccccssseseceeeceaseeeeeeeaneesees Derived StatisticS taisuri ie Taa a eaea ad a S Be eee eon ae aarAR HOD HH na xi Contents Section 3 TeamQuest Baseline Administration 3 1 tani Elemen eiea e a ielledes 3 2 Aggregation Sets esccecesceseeeeeeeeeeeeeeeeeeeeeeeesenaeeeteneeess 3 3 Alarm Handling serere e a T a Ra 3 3 1 Default Threshold Definitions 0 00 cece 3 3 2 Using the Default Threshold Definitions _ 3 3 3 Changing the Default Threshold Definitions 3 4 Whatis N xt castes catichi ETATE iinet aki Part 2 Customizing Your Site Section 4 Section 5 xii TeamQuest Performance Software Administration 4 1 INMTOGUCTION serata a a a a A 4 2 What is a Session nssesseeesisesirsssresrrssrnssrrrsrrrssre 4 3 Activating TeamQuest Performance Software A
4. Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec run Throughput runs hour Population open runs Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec run Throughput runs hour Population open runs System Log Overall Deadline WLS1 WLS K words of memory per deadline batch run for XPA systems Deadline batch tape I Os per second Deadline batch disk I Os per second Deadline batch I O SUPs per second Deadline batch ER CC SUPs per second Deadline batch IP SUPs per second Seconds per deadline batch run Deadline batch runs per hour Average deadline batch runs open System Log Workload Deadline WLS1 WLS all K words of memory per deadline batch run for XPA systems Deadline batch tape I Os per second Deadline batch disk I Os per second Deadline batch I O SUPs per second Deadline batch ER CC SUPs per second Deadline batch IP SUPs per second Seconds per deadline batch Deadline batch runs per hour Average deadline batch runs open TQ 01011 18 OS 2200 Parameters Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs
5. If a text string is used it must be enclosed in double quotation marks Export COMMA SPACE TAB text TQ 01011 18 Prints only the data All other header footer and title information if left off the resulting report In addition the column headings are printed only once and all page sizes are ignored Use this key for exporting data that is intended to be imported into a spreadsheet or a database package The Export key specifies the delimiter to be used between fields in the report The delimiter can be one of the reserved words TAB SPACE or COMMA or it can be any text string If a text string is used it must be enclosed in double quotation marks The Export key can be overridden by the x option on the tqrep command line Host Based Reporting for Table Data E 3 2 Table Section Keys The keys listed in this section are valid in the Table section of a report definition file The key names are not case sensitive Unless otherwise noted the key names can be specified only once If a key is specified more than once the last key value specified is used TableName class name The name of the table to be reported on This key is required in a table section For example in the table name Alarm Alarm Log Alarm is the class and Alarm Log is the name If the table name contains a space as in the example Alarm Log it must be enclosed in double quotation marks If the table name does not contain any
6. asg t t pr brkpt print t pr mfdrpt account_summary eof brkpt print ed uq t pr F Account 3 on dsplit split tpf a D 20 i time 86400 system eof filename tqtblprb f t pr t 2200 accountdisk end By placing an RUN card with a start time as the first image and an START card as the last image this can be made into a runstream that can execute every day Add an START image to your system s AUTO START element so the runstream will be placed in backlog Note The interval should be given in seconds The value of 86400 is the number of seconds in 24 hours 9 58 TQ 01011 18 Customizing Data Collection Probes Example 2 A little more sophisticated example follows In this example you only want to collect disk space usage by owner statistics when the amount of available fixed mass storage falls below 10 percent First you need to create a threshold to check for the TeamQuest Probes tqgmsp statistic system MassStor by System LOCAL In Use to be greater than 90 percent See 7 4 on how to create a threshold Then you need to modify the example table specification to include a time type of Table_Time_Event and to remove the Interval field Next you need to modify the data collection runstream by removing the interval Finally you would add user code to USRALARM C element that will issue an START of the data collection runstream compile USRALARM C and re link tqalm See 7 19 on how to ad
7. Appendix I 2200 Socket Interface 1 1 Communications Application Program Interface _ 2 1 2 TOTOE ieren dh tat kata iain tate I 2 1 2 1 Background RUN un eeeeeceeeeeeeeeeeeeeeeneeeeeeennaeeeenes I 3 1 2 2 Processor Gall simoniaa geiena iatea l 4 1 2 3 Configuration Element ccccseeeeeereeeesteeeeeees 4 1 2 3 1 Same User id For All TQTCP Runs _ I 5 1 2 3 2 Unique User ids For Each TQTCP Run l 7 1 2 4 Error Messages srovei En I 9 1 2 5 Security Considerations ccecesseeeeeeeeeeeteeeeeees I 10 1 2 5 1 User id and Password Encryption I 10 1 3 Connecting to the Socket ceesceeeeeeeeeeeeeeteeeeeeeeneeeees I 11 xviii TQ 01011 18 Contents Appendix J 2200 SNMP Trap Alarms J 1 OVOGWICW R E aie Ad ieflealh Pa ee J 1 J 2 CMS 1100 Configuration o eceeeceeeeeeeeeeeseeeeeteeeeeeeeeees J 2 J 3 TeamQuest Alarm Administration Interface 0 J 2 J 4 SNMP Trap Contents ccccceeseceeeseeeeeeeneeeeeeeeeeeeeeees J 4 J 4 1 SNMP Trap Mappings ccccccecessceeeeteeeesseeeeeees J 4 J 4 2 Variable Fields seanina ri ARA EEr J 5 J 5 Management Console Configurations eeeee J 6 J 5 1 HP OpenView Configuration eecceeeeeeeerees J 7 J 5 2 Tivoli Management Environment Configuration J 7 J 5 3 Computer Associates Unicenter TNG nanese J 9 J 6 TeamQuest MIB Definition oo eeeeeeeteeeeeeeeeeeeteeeeeee
8. Control Unit Disk TQ 01011 18 Subcategory Utilization Collection Base Hardware Interrupts Utilization Collection Base Hardware Interrupts Subcategory Total TIP Mem Trans Abs TIP Mem Trans ZOOMs Non TIP Mem Trans Abs Non TIP Mem Trans ZOOMs TIP Trans Abs TIP Trans ZOOMs Demand Abs Demand ZOOMs Batch Abs Batch ZOOMs Real Space Usage Working Set Mgmt Working Sets Subcategory Summary Summary Cache Summary Read Hit Write Hit Read Miss Write Miss Read Bypass Write Bypass Store Through Summary Summary A 3 OS 2200 Parameters Category Group Queue Tape Other Cache Disk Record Lock Processor Memory File XPC XIIP Mulithost Category System Return Available Summary Summary Read Hit Write Hit Read Miss Write Miss Read Bypass Write Bypass Store Through Summary Read Hit Write Hit Read Miss Write Miss Destage Bypass Misc Exec Summary Utilization Summary TIP UDS Summary UDS RLP Locks UDS RLP Unlocks UDS RLP Misc Subcategory Entries Server MaxCount Entries MaxCount TQ 01011 18 OS 2200 Parameters A 2 1 System Statistics The following statistics are the general statistics classified in the System category group Parameter Hierarchy Category Group Category Statistic Name Total Runs Open Demand Runs Open Batch Runs Open Demand Runs In Core Batch Runs In Core Backlogged Runs TIP Schd Queue Size Category Group Cat
9. E 4 Using Default Reports Each table in your performance database can have one or more default reports defined for it To see a list of the default reports available for a table use the l option of tqrep You can use the default reports in the following ways e Generate your report using one of the default reports The n option of tqrep allows you to specify the name of the report you want to generate e Place a copy of one of the default report definitions in a file by using the z option You can edit this file to create a new report definition Then use the new report definition file as input to tqrep by using the f option E 5 Exporting Data The tqrep program allows you to export data rather than printing a formatted report The resulting output file can be used as input to a spreadsheet program To export data enter the following command tqrep f alarm rep x comma end The resulting output file would look like this estamp Severity Alarm Id Value Qual 1 Qual 2 Qual 3 04 10 14 CRITICAL 4usrcpu 2 53 04 10 21 CRITICAL syscpu 15 28 0 23 NORMAL syscpu 1 75 0 24 CRITICAL 4syscpu 5 42 0 25 NORMAL syscpu 4 38 04 10 26 CRITICAL syscpu 7 23 0 0 0 30 NORMAL syscpu 4 92 gt 51 CRITICAL 4syscpu 7 65 57 NORMAL 4syscpu 1 33 GO OGG OGOGO SO 4H oO BK E 12 TQ 01011 18 Appendix F The tqgetm Program for TeamQuest Model The tqgetm pr
10. Metric Can be either ON or OFF ON Displays a single character metric prefix character instead of scientific notation when the string representation of values exceeds 1000 Numbers outside the range of 1e27 and 1e 27 are still displayed as scientific notation OFF Displays scientific notation when the size of the displayed number exceeds six digits This is the default if not specified Ideally the difference between the values specified in the minimum and the maximum fields should be an integer multiple of the value specified in the Major field If values are not specified for the Minimum and Maximum fields TeamQuest View uses the autoscale feature to present parameter values XAxis Style Font Style Size Foreground Background Border The specifications for the text attributes of the title of the horizontal axis Font Specifies the typeface in which the x axis title is printed Style Specifies the kind of emphasis for the typeface of the x axis title Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the x axis title be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Defines the number of points for the typeface of the x axis title This value determines the height of
11. Run_ID VEN AND Account 102 0 9 9 AND Account 10229 OR PID 111 OR PID gt 130 AND PID lt 135 This workload contains information for the following e Run VENCTL e Runs that begin with VEN and have account numbers 10209 10219 and 10239 through 10299 e PID 111 and 130 through 135 TQ 01011 18 5 7 Customizing Workloads 5 3 Maintaining Workloads with TeamQuest Performance Software Administration To perform any administration tasks with TeamQuest Performance Software Administration you must have a session established to the host software see 4 2 You access the dialog boxes for performing administration tasks by selecting the Administration menu Note When maintenance tasks are being performed the database is locked Only the user performing the tasks can access the database for modifying or updating workloads Other users can view workload information in read only mode when the database is locked TeamQuest Performance Software Administration allows you to perform the following tasks Performing Workload Administration Tasks see 5 3 1 Creating a New Workload Set Definition see 5 3 2 Displaying or Modifying a Workload Set Definition see 5 3 3 Deleting a Workload Set see 5 3 4 Creating a New Workload see 5 3 5 Displaying or Modifying a Workload Definition see 5 3 6 Importing Workload Sets and Workload Definitions see 5 3 7 Exporting Workload Sets and Workload Definitions se
12. You can use two local variables to make date time and system specifications easier These variables are time supplies the current date and time system supplies the system name from the s option if specified If the s option is not specified it supplies the system name from the program s initial register set Using these variables the group key values become time 1800 system B This line is the first record for the sample Record value lines contain all fields of a record for the sample interval This line consists of the Account of Fixed Files and Disk Space fields from the table specification The value for the Account field is 123456 the value for of Fixed Files is 7 and the value for the Disk Space field is 20345 Additional records follow on subsequent lines c This line is the required blank line that separates data samples TQ 01011 18 9 57 Customizing Data Collection Probes When creating input for data storage the following apply e All field values must be separated by spaces e Ifthe value itself contains spaces it must be enclosed in double quotation marks e Values must appear on the line in the same order as the fields are specified in the table specification 9 11 2 3 Examples of Table Data Collection The examples in this section show how to collect and store data into the example table Example 1 The commands in the following addstream collect disk space usage by account
13. tq2200AlarmTrap 9 tq2200AlarmDatabase OBJECT TYPE SYNTAX DisplayString SIZE 0 256 ACCESS read only STATUS mandatory DESCRIPTION The current database name tq2200AlarmTrap 10 tq2200AlarmParameter OBJECT TYPE SYNTAX DisplayString SIZE 0 256 ACCESS read only STATUS mandatory DESCRIPTION The current parameter string tq2200AlarmTrap 11 END TQ 01011 18 J 11 2200 SNMP Trap Alarms J 7 J 12 Special Considerations The TeamQuest Performance Software sends SNMP traps to the CMS 1100 SNMP agent The CMS agent actually sends the traps to the network managers The trap destinations are configured in CMS 1100 You specify the community name and host addresses where the trap is to be sent The SNMP trap port number used by CMS 1100 is always 162 There is no way to change this port number in CMS 1100 The length of the trap variable bindings is limited to 400 bytes by CMS 1100 the limit is 128 bytes prior to CMS 8R3A To stay within the byte limit variable fields may be removed from the trap by the TeamQuest Performance Software before the trap is sent to CMS 1100 The TeamQuest Performance Software runs on a variety of UNIX Linux platforms Microsoft Windows systems and OS 2200 systems Due to CMS 1100 limitations the TeamQuest SNMP traps on OS 2200 systems use a different enterprise identifier and different specific trap values than the UNIX and Microsoft Windows systems TeamQuest traps I
14. Both the Communications Application Program Interface subsystem and background run are automatically installed using SOLAR You must also create a data file called COMAPI COMNFIG that contains the appropriate Communications Application Program Interface configuration entries as described in the ClearPath Enterprise Servers Communications Application Program Interface COMAPI User s Guide In addition a PROCESS statement for Communications Application Program Interface must be added to your CMS 1100 or CPComm configurations The Communications Application Program Interface background run must be active on your system whenever you are using a socket connection to your OS 2200 system through TeamQuest View or TeamQuest Performance Software Administration For more information on how to install and configure the Communications Application Program Interface update your CMS 1100 or CPComm configuration for the Communications Application Program Interface and manage the Communications Application Program Interface background run see the ClearPath Enterprise Servers Communications Application Program Interface COMAPI User s Guide that accompanies your ClearPath OS 2200 release documents TQTCP The TeamQuest Baseline includes a TQTCP executable and background run The TQTCP background run uses the Communications Application Program Interface to listen for incoming connections from TeamQuest View or TeamQuest Performance Software Administration
15. Index tqtpm 1 9 9 17 tqusrprb 1 10 9 38 processor call for TQTCP I 4 processor reports B 6 B 8 product capabilities 1 5 program load statistics A 18 Q qualifying parameters 8 12 queue statistics A 39 R recovery of performance database 10 8 registering tables 11 2 11 18 removing aggregation set 6 19 reorganizing table files 11 26 report definition directory B 1 description of predefined reports B 1 history 1 11 host based D 1 monitor 1 11 user defined 1 11 report definition file description C 2 example C 1 for tqrep program E 3 format C 2 general section C 3 notation conventions for C 2 parameter group section C 29 resource 2 2 retention periods for tables 11 2 Retention Specification Menu of tqzip 11 9 retention specifications adding 11 13 deleting 11 14 description 11 9 listing 11 11 modifying the primary retention 11 12 setting start time 11 15 Runs Table System Activity Monitor A 58 TQ 01011 18 S saving derived statistic 8 13 security considerations for TQTCP I 10 sending SNMP traps J 1 Server table information A 62 Server Table Open Distributed Transacton Processing A 62 Service Table Open Distributed Transaction Processing A 64 session timeout messages 4 7 sessions named pipe connection 4 6 socket connection 4 5 SSL connection 4 6 sessions socket connection 4 1 setting retention start time 11 15 Single Point Operations SPO considerations for using wi
16. The following example shows how to list the aggregation sets for the default database gt tqasa gt TOASA 7R4G 011312 11 04 45 Fri Jan 13 16 22 10 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt Aggregation Set List gt gt Max gt File gt Name Gran Start Stop Keep Size Source Name gt SSAA Sele eo wees Sys a aaa a A a a a a aaa SoS Sieve aiaa Aa A Seve a gt 1l minute 00 01 00 00 00 00 23 59 59 08 00 00 23034 Trk gt 10 minute 00 10 00 00 00 00 23 59 59 8 days 13818 Trk gt l hour 01 00 00 00 00 00 23 59 59 35 days 8396 Trk 10 minute 8 hour 08 00 00 00 00 00 23 59 59 400 days 11996 Trk 1 hour gt gt An following Name indicates an expanded aggregation set gt gt gt End TQASA TQ 01011 18 Customizing Aggregation Sets 6 4 4 Unloading Aggregation Set Definitions The tqasa program allows you to export all of the aggregation set definitions currently contained in the performance database into a file You can use this file to view the aggregation set definitions and update them by using a text editor You can then reload the aggregation set definitions by using the l option on the tqasa processor call see 6 4 5 To unload aggregation set definitions with tqasa use the following processor call format tqasa d databasename u aggset file where d specifies the database for which you want to unload aggregation set definitions The databa
17. The performance database is made up of a set of Exec files The file TQDATA DBNAMES is common to all performance databases This file is created and initialized when the first performance database is created This file is deleted when the last database is removed In addition the following files are created for each database TQDATA xxxxMISC TQDATA xxxxCONFIG TQDATA xxxxDICTION TQDATA xxxxWORK TQDATA xxxxTHRESHLD TQDATA xxxxLOCK e TQDATA xxxxaaaaaaaa contains the data for an aggregation set defined for the database aaaaaaaa is the first eight characters of the aggregation set name Because the aggregation set name is used in the file name it must be unique within the first 8 characters Also aggregation set names can only contain the characters a z A Z 0 9 and hyphens e TQDATA xxxxALARMLOG is created if it does not exist when executing tqalm e TQDATA xxxxODTPFILE is created if it does not exist when executing tqodtp TQDATA xxxxSAMFILE is created if it does not exist when executing tqsam e TQDATA xxxxTLIST and TQDATA xxxxINTSYSTB are created if they do not exist when executing tqalm In the preceding file names xxxx is the first four characters of the database name All TeamQuest Baseline and TeamQuest Online files have a qualifier of TQDATA When you are creating a database if a database file already exists it is initialized If you want your performance database files to be catalog
18. The tqusrprb probe can be started in continuous probe by supplying the m option When running in continous mode the tqusrprb program processes multiple cycles of PRBDATA as they become available determined by the ability to exclusively assign each cycle Your program that you supply to create the PRBDATA file must assign the next cycle by ASG CP PRBDATA 1 writing data to it then doing an FREE PRBDATA 1 When the tqusrprb probe can exclusively assign that cycle the tqusrprb probe then stores the data from that cycle and either processes the next cycle or waits until the next cycle becomes available The tqusrprb probe continues to run until the console keyin TQUSR END is entered or keyin name END is entered where keyin name is the name supplied through the k option You may start several tqusrprb probes to continuously run if you have different types of data to store from different sources such as two different formats of TQARG and PRBDATA This can be done by placing USE names on TQARG and PRBDATA prior to calling the tqusrprb probe or by changing the qualifier on the runstream Since keyin names must be unique you must supply a unique keyin name through the k option for runs of tqusrprb other than the initial run which uses the default keyin name of TQUSR You stop each run of tqusrprb by issuing the associated END keyin name used when tqusrprb was started The processor call format of tqusrprb is sys lib tq baseli
19. W l LELPPPpPpPIP aRBRW OONDONMND v y vuU te ON xvii Contents Appendix C Report Definition File C 1 IATLOGUEGHION Avcdesiteiearesihete aidicdiseenensaies C 1 C 2 Notation Conventions cccccccseseeceecceeceeeeeeeeeeeeeneneees C 2 C 3 File Foma cchveceteaacied e ae a e Aa tat at C 2 C 4 General Section eeceeececccccccccccceeeeceseeaeaeaeeeaeeseseceeeeeeeess C 3 C 5 Parameter Group Sections eceeseeeeeeeeesteeeeeteeeeees C 29 C 6 Individual Parameter Setting Sections ceeee C 32 Appendix D Host Based Reporting for Aggregation Set Data D 1 Using tqrpt for Condensed Output eee D 4 D 2 Using tqrpt for Reporting Derived Statistics D 5 Appendix E Host Based Reporting for Table Data E 1 Overview of tqrep sesssssssssssssssresssessresssrnssrenssrnnssrnnernnent E 1 E 2 Command Format ccessceeeeeeeeeeeeeeeeceeeeseeeeseneeeeeneeeess E 2 E 3 Report Definition File Format oo eeeeceeeeeeseeeeeeeeeeee E 3 E 3 1 Report Section KeyS oo eeesseeeeeeeeeeeteeeeeteeeeeseeees E 5 E 3 2 Table Section Keys eccccesceeeeeeeeeeeeeeeeeteeeeeneeeeees E 8 E 3 3 Using Special Identifiers in Section Keys _ E 10 E 4 Using Default Reports oo ee eeceeeeeeeeeeneeeeeeeeneeeeetee E 12 E 5 Expono Daa since ina Anais Appendix F The tqgetm Program for TeamQuest Model Appendix G Alarm History Table Appendix H tqzip Processor Call Format
20. e System specifies the system on which the data was collected A system is defined as a group of closely coupled components that function together and allow measurement of system performance characteristics e Category group identifies a group of categories with similar characteristics Any given system may have many categories and the category group specification is used to group related categories For example category groups can include processor memory IO activity and system e A category breaks down a category group Examples of categories within the category group memory are utilization dynamic allocator and program load e A subcategory further breaks down a category into a group of related statistics For example a category called processor can be grouped into subcategories of Exec requests utilization and hardware interrupts e A statistic name is the name of the information being collected and stored in the database For example percent busy requests per second and K words per request are statistic names The following examples show the hierarchy of key names for various statistics System Category Group Category Subcategory Statistic Name CLO1 Processor Total Utilization Total Exec CLO1 IO Activity Logical Channel CacheSummary Transfer Time CLO1 Memory Paging Activity Working Sets Inputs Sec TQ 01011 18 2 2 3 TeamQuest Baseline and TeamQuest Online Concepts Resource Specification A parameter can be fu
21. 4 TQ 01011 18 Glossary SOLAR See Software Library Administrator system log A file maintained by the Exec to keep historical record of the activity in the system The system log stores for example the time a run starts and ends console messages checkpoints and file information T TeamQuest Model A software product that lets you build and solve a network queuing model and thus predict the performance of your computer system The tqgetm program of TeamQuest Baseline retrieves performance data that can be input to the OS 2200 Model Builder Wizard TeamQuest Performance Management Software An open architecture designed around the client server model and upon which all products in the TeamQuest Performance Management Software family are based TeamQuest View The workstation software that provides the graphical user interface to the host software TeamQuest View is the client in the client server model of the TeamQuest Performance Management Software family TIP Performance Monitor TPM TPM provides performance analysis of TIP HVTIP transactions trend analysis The process of identifying periodic oscillations continuous growth and continuous decline for the workloads on a computer system W workload The logical classification of work on a computer system TQ 01011 18 Glossary 5 Glossary X XIIP The XPC interface that connects an XPC to the system One or more XIIPs can be used per XPC XPA Refers
22. At the prompt enter the resource for the statistic You can enter either the number of the resource or the resource name When the parameter has been selected the tqzip program displays the full name of the parameter It then displays a series of threshold definition prompts You can enable or disable individual threshold definitions Enter y at the prompt to enable threshold checking for the parameter Enter n at the prompt to disable threshold checking for the parameter You can create a threshold definition for a parameter even if you enter an n at this prompt The threshold definition will exist in the performance database but will not be used until you update it to enable threshold checking by entering a y at the prompt Enter the number of threshold zones for the parameter You can define either 2 or 3 threshold zones for a parameter See 2 5 for more information on threshold zones Enter the number of the normal zone for a sampled value for this parameter This is used to determine if a zone change has occurred on the first sample taken by a probe Enter the first threshold value for the parameter This value is used to determine the boundary between threshold zone 1 and threshold zone 2 If you are defining three threshold zones enter the second threshold value for the parameter This value is used to determine the boundary between threshold zone 2 and threshold zone 3 Specify the number of seconds that the current value
23. Customizing Data Collection Probes 9 11 2 Writing a User Probe for Table Data 9 48 The tqtblprb program provides a method of creating populating and removing TeamQuest Baseline tables The program is intended for use with tabular data that you would like to collect and store for analysis with the TeamQuest View user interface The tqtblprb program reads and processes ASCII input that can be produced by SSG skeletons IPF procedures or by programs written in any programming language This section explains the two step process of creating and then populating a TeamQuest Baseline table Throughout this section you will find references to a specific example which will provide a clearer understanding of the concepts and procedures involved in this process Example At the end of each business day you want to know how much fixed mass storage disk space is being used by each account Special Syntax When creating input for tqtblprb follow these general rules e You can use the backslash character to continue input to the next line e Items on a line are separated by spaces or tabs e Items that contain spaces or special characters must be enclosed in double quotation marks e An item enclosed in quotation marks cannot be continued to the next input line TQ 01011 18 9 11 2 1 Customizing Data Collection Probes Creating a Table Before tabular data can be stored into the performance database a table must ex
24. For example you can specify that data is sampled from 08 00 to 15 59 or that it is sampled for a complete 24 hour interval from 00 00 to 23 59 From this time interval TeamQuest Baseline calculates the number of time groupings within each type If you collect data for 24 hours each day and the data has a granularity of 30 minutes you will have 48 time groupings for each day Figure 6 1 shows the logical layout of how the data records for a parameter are grouped in an aggregation set The aggregation set in the figure is defined to collect and store data samples at 30 minute intervals daily and retain that data in the database for a period of 60 days Data is collected for a 24 hour period So for this example the granularity is 30 minutes the type of date consolidation is day the retention period is 60 days the start time is 00 00 00 the end time is 23 59 59 Only one parameter Parameter N is shown in this figure However the aggregation set would contain separate groupings for each parameter collected and stored TQ 01011 18 TQ 01011 18 Parameter N Figure 6 1 Customizing Aggregation Sets Date Grouping Day 1 Date Grouping Day 2 Date Grouping Day 60 Time Grouping 30 minute interval 1 30 minute interval 2 30 minute interval 3 30 minute interval 48 Time Grouping 30 minute interval 1 30 minute interval 2 30 minute interval 3 30 minute interval 48 Time Grouping 30 minute interva
25. If this parameter is not specified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name specifies the aggregation set to be used for data collection The aggset1 and aggset2 parameters are the names of aggregation sets defined in the database name The aggset1 and aggset2 parameters are case sensitive All aggregation sets are optional on the call line If aggregation sets are not specified the Aggset1 and Aggset2 settings of the Aggregation Sets section in the tq ini element are used See 2 2 1 for a description of aggregation sets specifies which audit trail number to use This option should only be used to read MIPS data from a trail other than the dedicated COD audit trail To read from the system log file use n audit trail number specification where audit trail number is the system log trail number Processing from the COD audit trail is much more efficient than using the n option Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent lines of the processor call The parameters specified on the processor call statement must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the
26. Pages of real memory used by users of TIP in pages Pages of real memory used by users shared in pages TQ 01011 18 OS 2200 Parameters Demand Batch Category Group Category Subcategory Statistic Name Imm Page Faults Sec Delayed Page Faults Sec Delayed with IO Sec Avg IO Time per Delay Avg Size User Queue Category Group Category Subcategory Statistic Name Inputs Sec Pages Input Sec Outputs Sec Pages Output Sec IO Time per Output TQ 01011 18 Pages of real memory used by users of demand in pages Pages of real memory used by users of batch in pages Memory Paging Activity Working Set Mgmt Page faults that did not deactivate the faulting process pages per second Delayed page faults where the faulting process was suspended pages per second Delayed page faults that had I O pages per second I O time per delayed page fault milliseconds per page fault Average number of pages on the page load request queue Memory Paging Activity Working Sets Working sets input working sets per second Pages of working set input pages per second Working sets output working sets per second Pages of working sets output pages per second Average I O time per working set output millisecond per working set A 23 OS 2200 Parameters A 2 5 A 24 I O Activity Statistics The I O Activity category group provides general statistics for I O activity Parameter Hierarchy Category Grou
27. Severity DataFields Alarm_ID LJ12 Alarm Id DataFields Parameter_Value JHHHHE 44 Value DataFields Qualififer_l LJ10 Qual 1 DataFields Qualififer_2 LJ10 Qual 2 DataFields Qualififer_3 LJ10 Qual 3 Sort Timestamp Ascending The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A This is the first line of the report definition file and the beginning of the Report section The section name must be enclosed in brackets and is not case sensitive Only one Report section can appear in a report definition file The Report section contains information that is applied to all the tables in the report The Report section defines such things as the name for the output file the time range of the data the time consolidation the page size of the report the report headings and footers and the report column delimiters The Report section name is followed by section keys with their assigned values This line begins the Report section keys See E 3 1 for a description of the valid keys for the Report section This line begins a Table section The section name must be enclosed in brackets and is not case sensitive A report definition can contain one or more table sections Each table section specifies the name of the table and the fields from the table that should be reported on The Table
28. TQ 01011 18 4 1 TeamQuest Performance Software Administration 4 3 4 2 Activating TeamQuest Performance Software Administration When TeamQuest Performance Software Administration has been correctly installed on your system you can activate the interface from the Start menu Select Start gt Programs gt TeamQuest gt TeamQuest Performance Software Administration TeamQuest Performance Software Administration is activated and you and you are prompted to create a session see 4 4 Once the session is established the TeamQuest Performance Software Administration window is displayed 4 TeamQuest Performance Software Administration DER File Edit Administration Help TeamQuest Copyright 2006 2012 TeamO uest Corporation All administration tasks are performed using the Administration menu in this window TeamQuest Performance Software Administration allows you to perform the following tasks e Maintaining Workloads with TeamQuest Performance Software Administration see 5 3 e Maintaining Aggregation Sets with TeamQuest Performance Software Administration see 6 3 e Maintaining Derived Statistics with TeamQuest Performance Software Administration see 8 3 TQ 01011 18 TeamQuest Performance Software Administration 4 4 Creating a Session A session defines how the TeamQuest Performance Software Administration connects to the host where a TeamQuest performance database exists The session inform
29. The following statistics are collected by workload set workload category group category subcategory and statistic Parameter Hierarchy Category Group Category Subcategory Statistic Name Requests Read Requests Words Regq In Words Req Out Cached Requests RET Hardware Serv Time Transfer Time Category Group Category Resource Subcategory Statistic Name Requests Read Requests Words Regq In Words Req Out Cached Requests RET Hardware Serv Time Transfer Time IOTRACE Workload Total Number of I O requests I O requests that were reads Words per input request Words per output request Number of I O request that were cached Time required by all system devices to process the I O file lock delay chan deviceQ hardware serv time Time required by the device to process the I O seek latency CU delay transfer time Time required to transfer the data IOTRACE Workload LCO LC1 Logical Channel Number of I O requests I O requests that were reads Words per input request Words per output request Number of I O request that were cached Time required by all system devices to process the I O file lock delay chan deviceQ hardware serv time Time required by the device to process the I O seek latency CU delay transfer time Time required to transfer the data TQ 01011 18 OS 2200 Parameters A 6 MAPPER Log Parameters This subsection contains a list of the parame
30. Valid best fit line types are None No best fit line Linear A linear best fit line Polynomial A best fit polynomial curve of any order form 2 through 5 Order is determined by the order value C 8 TQ 01011 18 Report Definition File Order The order of the polynomial when using the polynomial line type Valid values are 2 3 4 or 5 Equation The setting to determine if the best fit line equation is displayed The text attributes are defined in the BF Note key Valid values are ON Display the best fit equation OFF Do not display the best fit equation XAxis Title text The title for the horizontal axis of the report It appears under the x axis of the report text Can be a string of text up to 80 characters long If it contains spaces the title must be enclosed in quotation marks Value Title text The title for the vertical axis of the report It appears along the value axis of the report text Can be a string of text up to 80 characters long If the title contains spaces it must be enclosed in quotation marks Title Style Font Style Size Foreground Background Border The specifications for the text attributes of the page title of the report Font Specifies the typeface in which the page title is printed on the report Style Specifies the kind of emphasis for the typeface of the page title Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emp
31. aggregation set specified as the a parameter on the tqbsp processor call statement see 9 3 1 TQ 01011 18 9 4 9 4 1 Customizing Data Collection Probes System Activity Monitor Collection Note The System Activity Monitor tqsam program is released with TeamQuest Baseline only It is not available with TeamQuest Online TeamQuest Baseline contains a system activity monitor runstream called tqsam The tqsam runstream uses a run identifier of tqgsam and executes the tqsam program The tqsam program collects the most current information available on system run and transaction activity on your OS 2200 system The tqsam program makes use of a set of Exec callable interface routines that return the most up to date information on all active runs and transactions on your OS 2200 system System activity is monitored by the tqsam program at every sample interval and data is stored in several tables Information on all active runs is stored in the Runs table and transaction information is optionally collected and stored in the Transactions table See Sections 2 and 11 for more information on the collection of table data tqsam Processor Call Statement The tqsam runstream contains the tqsam processor call statement that controls the system activity monitor routines to collect information on all active runs and the longest running transactions on your system The tqsam processor call statement includes options to collect statistics on
32. for calculating the weight Non_Seq_Cons_Method cons method field name Specifies a nonsequential consolidation method for the field Non_Seq_Cons_Method is normally used when combining records within the same time period Non_Seq_Cons_Method is the keyword for the item cons method field name 9 54 can be one of the following See 9 11 2 5 for a description of these consolidation methods Id_Cons Default Min_Cons Max_Cons Sum_Cons Weight_Cons First_Cons Last_Cons is valid only for Weight_Cons types and is the field to use for calculating the weight TQ 01011 18 Customizing Data Collection Probes Example In the example the first field definition is Time Time is a group key and a primary key The field type is a timestamp and is specified as Field_Type_Timestamp The data subtype is time and date and is specified as Time_Date_Subtype The field is consolidated by the last value This field definition is as follows Field Field_Name Time Group_Key Primary_Key Field_Type_Timestamp Time_Date_Subtype Seq_Cons_Method Last_Cons Note This field definition is for the sample time and should be the first field in every table Creating the Example Table The following commands will create the table specification input file TPF dsuprobe tab ed iq dsuprobe tab TABLE Table_Name 2200 accountdisk A File Name usertabs eld_Name Time B roup_Key rimary_Key eld_Type_Timestamp m
33. provides information about the available tables in the database Included in the information for each table is a list of the default report names for the table and a list of the field names and type of data contained in the field specifies the start time for the data The starttime parameter is in the form of MMDDYYYY HHMMSS If no value is specified the StartTime value from the report definition is used specifies the end time for the data The endtime parameter is in the form of MMDDYYYY HHMMSS If no value is specified the EndTime value from the report definition is used specifies the time consolidation for the selected records The time consolidation parameter has the format of nnHOUR nnMIN or nnSEC where nn is the number of units If no value is specified the TimeCons value from the report definition is used specifies the name of the table to report on The tablename parameter is expressed as class name For example in the table name Alarm Alarm Log Alarm is the class and Alarm Log is the name If the table name contains a space as in the example Alarm Log it must be enclosed in double quotation marks If the table name does not contain any spaces the quotation marks are not needed prints the default report for the table specified by the t option If you use the r option you must specify the t option prints the report definition of the first default report for the table specified by the t option If you use
34. set one day prior to the last point it had previously consolidated It continues to consolidate data until it reaches the end of the data in the source aggregation set Using the r option allows you to repopulate an aggregation set after a change in the content of the source aggregation set This may be necessary if data is loaded into the database for a time period that had already been processed by tqagg The default action of tqagg is to reprocess the data points from the previous day at each execution This is necessary to ensure that the data collected by the probes running in offline mode is consolidated that is tqsys tqgmpr tqtlp tqtpm tqrtpm and tqiot If tqagg starts before the offline probes are finished for the day some of the data stored by the offline probes would not be aggregated until the next day when tqagg is run If you do not use any of the offline probes or can ensure that tqagg does not run until after all of the offline probes are complete you can inhibit any reprocessing of data by specifying the r option with 0 days When 0 days is specified tqagg starts aggregating the data immediately following the last point it had consolidated allows you to set the source aggregation set allows you to set the destination aggregation set e The r option is only required in special cases Normally you should not need to use the r option e By default the SYS SLIB RUN TQAGG runstream is started automatically
35. specifies the end time for the data The endtime parameter is in the form of hhmmss The default is 235959 specifies the number of minutes to consolidate for each point The time consolidation parameter must be an integer This number will be rounded down to the nearest whole interval stored in the aggregation set Host Based Reporting for Aggregation Set Data f specifies the starting date for the data The firstdate parameter is expressed as mmddyyyy mmddyy or mmdd The current date is the default If a 2 digit year is specified the digits 00 through 27 are interpreted as the years 2000 through 2027 and the digits 28 through 99 are interpreted as the years 1928 through 1999 You can express the firstdate parameter as a date relative to the current date by entering n where n is the number of days before the current date l specifies the ending date for the data The lastdate parameter is expressed as mmddyyyy mmddyy or mmdd The current date is the default If a 2 digit year is specified the digits 00 through 27 are interpreted as the years 2000 through 2027 and the digits 28 through 99 are interpreted as the years 1928 through 1999 You can express the lastdate parameter as a date relative to the current date by entering n where n is the number of days before the current date p prints a list of stored and derived parameters and then exits the program W specifies the number of characters to print per line The defau
36. throughput response time lines processed per second and IOs processed per second B 21 Appendix C Report Definition File C 1 Introduction Each predefined report is defined in its own report definition file You can also define your own reports to meet the needs of your site Use the information in this appendix to customize reports for your site You can create new reports by either creating a new report in TeamQuest View and adding and deleting parameters or by creating your own reports using a text editor and entering the file format described in this appendix The following example shows a report definition file The name of the report definition file is dsk unit rpt and it defines the Activity by Disk Unit report general Copyright Copyright c 2012 TeamQuest Corporation All Rights Reserved Host Type 2200 Report Name Activity by Disk Unit Format Date Time System Application Group t wWorkload Set ZWorkload Category Group Category Subcategory Statistic Value Resource Parm0 System Category Group IO Activity Category Disk Subcategory Summary Statistic Reqs Sec KWords Req Percent Busy RET FLD Hardware Serv Time Transfer Time Concurrency Resource Value Types Average TQ 01011 18 C 1 Report Definition File C 2 Notation Conventions The following conventions apply to the format of a report definition file as descri
37. 10 Activity Disk I0s Sec gt 4 CLO1 10 Activity Disk Summary Percent Busy DSKF1l gt gt Enter number name lt blank gt for more or E to exit gt gt gt DERIVED Derived Derived Category Derived Subcategory UserCpuUtil B gt Delete this threshold y lt n gt gt y gt gt The threshold definition has been deleted gt gt Do you want to delete another threshold y lt n gt C gt gt gt Threshold checking is ON gt Notify active probes of these threshold changes y lt n gt D gt y TQ 01011 18 7 19 Customizing Alarms and Thresholds The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A 7 20 The tqzip program displays the list of threshold definitions currently defined in the performance database Select the one you want to delete by either entering the number or full name at the prompt Derived statistic names begin with DERIVED The tqzip program displays the name of the selected parameter or derived statistic for verification before the deletion is performed If you enter y at the prompt the threshold definition is deleted from the performance database If you enter n at the prompt the threshold definition is not deleted from the performance database If you want to delete another threshold definition enter y at the prompt The tqzip program returns you to the prompt that
38. 12 7 13 Customizing Alarms and Thresholds Notifying Probes of Threshold Changes The TeamQuest Baseline probes use the threshold definitions created by tqzip to determine the threshold checking to be done When a probe is initialized the threshold definitions are read from the database The probe uses this initial set of threshold definitions during its entire execution unless it is notified that threshold changes have been made To notify all active TeamQuest Baseline probes that changes have been made to threshold definitions use the Notify Probes of Threshold Changes selection on the Alarms and Thresholds Menu see 7 3 Enter 4 at the prompt and press the transmit key Upon being notified the probes re read the threshold definitions from the database and obtain the latest set of thresholds available You must notify the probes of any threshold changes after defining updating or deleting threshold definitions If you do not notify the probes of the threshold changes the changes do not take effect until the probes are reinitialized Unloading Threshold Definitions The tqzip program allows you to export all the threshold definitions currently contained in the performance database into a specified file You can then use this file to view all threshold definitions update the threshold definitions using a text editor and reload them at a later time by using the l and t options on the tqzip processor call see 7 14 To unload
39. 12 2 Executing tqrorg To execute the tqrorg utility use the following command line format tqrorg table filel table file2 table file3 where table file1 table file2 and table file3 are names of table files that exist in the performance database If you need to obtain the file name for a given table use the Tables Menu from the tqzip program and select List Tables see 11 4 If you attempt to reorganize a file that is not a table file an error message will be printed but no damage will occur to the specified file If you attempt to reorganize a file that is currently open by another run the reorganization will not be done In the event of a system crash the file open count can become incorrect Before the reorganization can take place the table manager subsystem needs to be deactivated by entering the control statement sys lib solar deact sys lib tq bslgate tq tmssd 11 28 TQ 01011 18 Appendix A OS 2200 Parameters A 1 Introduction This appendix contains a list of the parameters collected by the TeamQuest Baseline and TeamQuest Online probes The parameters listed in this appendix are classified by the hierarchy of key names The hierarchy can consist of the system category group category subcategory and statistic name This appendix lists the parameters collected by the SIP TPM RTPM MIPS IOTRACE MAPPER log system log and ODTP probes TQ 01011 18 SIP Parameters TPM Log Parameters MIPS Para
40. 30 minute interval 1 Date Grouping 1 hour interval 1 Day 1 30 minute interval 2 Week 1 1 hour interval 2 30 minute interval 3 1 hour interval 3 30 minute interval 48 1 hour interval 24 Parameter Parameter 1 1 Time Grouping Time Grouping Date Grouping 30 minute interval 1 Date Grouping 1 hour interval 1 Day 60 30 minute interval 2 Week 52 1 hour interval 2 30 minute interval 3 1 hour interval 3 30 minute interval 48 1 hour interval 24 Time Grouping Time Grouping Date Grouping 30 minute interval 1 Date Grouping 1 hour interval 1 Day 1 30 minute interval 2 Week 1 1 hour interval 2 30 minute interval 3 1 hour interval 3 30 minute interval 48 1 hour interval 24 Parameter Parameter X X Time Grouping Time Grouping Date Grouping 30 minute interval 1 Date Grouping 1 hour interval 1 Day 60 30 minute interval 2 Week 52 1 hour interval 2 30 minute interval 3 1 hour interval 3 30 minute interval 48 1 hour interval 24 Figure 6 2 Data Consolidation from Aggregation Set 1 to Aggregation Set 2 TQ 01011 18 6 2 4 6 2 5 Customizing Aggregation Sets Feeding Aggregation Sets When an aggregation set receives its data from another aggregation set the second aggregation set is said to be fed by the first aggregation set For example if data from Aggregation Set 1 is consolidated into Aggregation Set 2 Aggregation Set 1 is the source of data for Aggregation Set 2 Aggregation Set 2 is fed by Aggregation Set 1 The
41. 32 hours TQ 01011 18 2 1 TeamQuest Baseline and TeamQuest Online Concepts 2 2 2 2 2 TeamQuest Baseline allows multiple aggregation sets in your database So for example if you are interested in real time monitoring of your system you might have an aggregation set that collects and stores data at 30 second intervals and retains that data for a 1 day time period If you are interested in trend analysis of your system you might want to store data at half hour intervals for a 6 month period You can have both of these aggregation sets in your database as well as others that you defined to meet your particular needs TeamQuest Baseline is released with some default aggregation sets see 3 2 You can also define your own aggregation sets by using TeamQuest Performance Software Administration see 6 3 Parameters The term parameter refers to a performance data item in the database A parameter consists of a statistic and optional resource workload and application group qualifiers The term parameter applies only to performance data stored in the performance database aggregation sets it does not apply to table data A statistic describes the information being collected and stored in the database A statistic is defined by a hierarchy of key names which identify the statistic for retrieval from the database The key names used for the classification of statistics are system category group category subcategory and statistic name
42. A range of consecutive characters can be specified by using a hyphen to separate the first and last characters desired For example A M matches the characters A through M The hyphen loses this special meaning if it occurs first after an initial caret 4 if any or last in the string Run_ID USERL A E matches on USERA USERB USERC USERD or USERE e Ifa caret is the first character in the brackets the regular expression matches any character except one of those found in the brackets For example Run_ID USER 0 9 matches any character following USER but does not match 0 through 9 e regexpiregexp2 is a regular expression that matches regexp1 followed by regexp2 For example 0 9 A F matches any digit 0 through 9 and any character A through F Txn CSFLD GJ 2 5 matches on CSF followed by D through G and 2 through 5 e regexp is a regular expression that matches zero or more occurrences of the regular expression regexp that immediately precedes it For example A matches a null string or A AA AAA and so on Run_ID RO T matches on anything beginning with R then zero or any number of the character O and ending with T This matches RT ROT ROOT ROOOT and so on e regexp is a regular expression that matches one or more occurrences of the regular expression regexp that immediately precedes it For example A matches A AA AAA and so on Run_ID RO T matches on anything beginning with R then
43. Baseline These products allow you to perform the following tasks e Analyze and view system performance data using a real time monitor e Present reports simultaneously from multiple hosts e Perform correlation analysis between performance statistics e Produce hard copy output of reports using a printer or a plotter TeamQuest Baseline provides the additional capabilities to perform the following tasks e Characterize the workloads on the system e Build and maintain a database containing performance data on global system activity network activity and database activity performance data by workloads and performance data on process activity e Perform threshold checking on performance data and derived statistics and generate alarms when values exceed predefined thresholds e Consolidate data collected at a fine granularity into an aggregation set of coarser granularity e Collect reduce and store process data e Create your own user data probes e Generate character based reports on the host system 1 5 Processing Overview TeamQuest Baseline and TeamQuest Online processing consists of two phases data collection and data display These phases include the various activities involved in creating the performance database maintaining the database and using the data stored in the database for report generation and graphics displays TeamQuest Baseline and TeamQuest Online control the activities for data collection and work with TeamQues
44. Deleting a Derived Statistic Definition 8 4 To delete a derived statistic definition from a database select a derived statistic from the Derived Statistics list box in the Derived Statistics Maintenance dialog box see 8 3 1 When you click Delete all of the derived statistic definitions are searched for references to the derived statistic you want to delete If a reference is found a warning dialog box informs you that you are deleting a derived statistic that is referenced in another derived statistic You are asked if you still want to delete the statistic When the derived statistic definitions are searched and no references are found to the statistic you want to delete an alert box appears to verify that you have not inadvertently pressed the Delete button When you click Yes in this alert box the statistic is removed from the Derived Statistics list box in the Derived Statistics Maintenance dialog box To delete the derived statistic from the database you must click OK Note If you are deleting a derived statistic that also has a threshold definition the threshold is also deleted You can cancel the deletion of a derived statistic either by clicking the No button in the alert box or by clicking Cancel in the Derived Statistics Maintenance dialog box see 8 3 1 Maintaining Derived Statistics with tqzip The tqzip program allows you to perform additional tasks associated with maintaining the derived statistics Using tqz
45. Description File Name Description File Name Description File Name Description File Name Description TQ 01011 18 LC UNIT RPT Activity by LC Shows detailed I O information for the selected logical channels LC TOPN RPT Most Active Logical Channels Shows detailed I O information for the top n logical channels DSK UNIT RPT Activity by Disk Unit Gives detailed I O information for the selected disk units DSK TOPN RPT Most Active Disk Units Gives detailed I O information for the top n disk units CD SUM RPT Cache Summary Contains overall cache summary information CD LCPER RPT Cache LC Percent Shows the percentage of requests handled by each cache action for the selected logical channels PC UNIT RPT Physical Chan Activity Shows detailed I O information for the selected physical channels PC TOPN RPT Most Active Physical Channels Shows detailed I O information for the top bottom n physical channels IO TOT RPT Total IOs Second Shows total system IOs per second as well as the minimum and maximum for the time range Predefined Report Descriptions B 3 2 The following predefined report definitions provide for accessing the performance statistics gathered by the OS 2200 SIP probe for the XPC The reports are in the io subdirectory of the sip subdirectory File Name Description File Name Description File Name Description X
46. Exception The patterns that are used to display thresholds on the report The patterns indicated by these specifications appear in a small box next to the parameter names in the legend of the report Normal Is the pattern that appears on the report when the data value is above the lower threshold limit and below the upper threshold limit Exception Is the pattern that appears on the report when the data value is below the lower threshold limit or above the upper threshold limit Valid patterns for Normal and Exception are Solid For no pattern Horizontal For a horizontal line pattern within the object Vertical For a vertical line pattern within the object ForwardDiagonal For a line pattern of stripes on a forward diagonal within the object BackwardDiagonal For a line pattern of stripes on a backward diagonal within the object Cross For a line pattern of square shapes within the object DiagonalCross For a line pattern of diamond shapes within the object The default values for Threshold Pattern are defined in the View Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Window Area top left bottom right The specifications that define the initial area of the report window The initial window is defined in pixels top Specifies the pixel location for upper coordinate of the top left hand corner of the window left Specifies the pixel location for the left coor
47. F This line begins the aliases section that defines the aliases associated with the derived statistic Each part of the alias definition must appear on a line by itself The actual values must be enclosed in double quotation marks G This line specifies the alias name H The Y indicates that the alias is used in the formula for the derived statistic If the alias is not used in the formula an N appears I This line begins the parameter hierarchy used to define the alias Each key name in the hierarchy must be on a line by itself See 8 2 2 for a description of the SYSTEM WLS and WL references in the parameter hierarchy J This line begins the second alias definition associated with this derived statistic If additional aliases exist for a derived statistic they would be listed in this format K A blank line preceded by a pound sign ends the derived statistic definition Additional derived statistic definitions can follow You can add additional derived statistic definitions by duplicating the preceding lines and modifying the fields within the double quotation marks L The last line of the thresholds file must be the END DERIVED STATISTICS line 8 20 TQ 01011 18 Customizing Derived Statistics 8 4 2 Loading Derived Statistic Definitions You can reload the performance database with a new or updated set of derived statistic definitions by using the l and s options on the tqzip processor call If you upda
48. FIPS 140 2 compliant encrypted user id and password can also be created The TQENC program is used for encrypting the user id and password used in the TQTCP configuration element To use this program you must have the Unisys Cipher API product installed on your system Cipher API is governed by the Federal Information Processing Standards FIPS Publication 140 2 Security Requirements for Cryptographic Modules When using the TQENC program you are prompted for a user id and password to encrypt The TQENC program then displays the configuration statements that can be cut and pasted into the TQTCP configuration element described in I 2 3 The following shows an example of using the TQENC program gt TQENC gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt Enter a user id to encrypt user id lt none gt gt useridl gt gt Enter a password to encrypt password lt none gt gt passwordl gt gt Include the following lines in your TQTCP configuration element gt gt euser O0dal84c595a3acd48b669b1520c76005 017 Depswd 639692a250c9e94a383a2F952e449b7468ec TQ 01011 18 2200 Socket Interface 1 3 Connecting to the Socket To access your OS 2200 system using a socket connection you must use TeamQuest View or TeamQuest Performance Software Administration The default port for socket connections is 2700 After you launch TeamQuest View or TeamQuest Performance Software Administration
49. In situations where 3 zones are defined the zone difference is either 0 or 1 or 2 If the difference is 1 and the normal zone is 1 or 3 the severity is WARNING If the difference is 1 and the normal zone is 2 the severity is CRITICAL If the difference is 2 and the normal zone is 1 or 3 the severity is CRITICAL Value This is set to an ASCII representation of the value that triggered the alarm ThreshZones This is set to a character string that represents the threshold s zones and the normal current and previous positions within those zones The following is the format of this string lt threshl gt lt thresh2 gt where lt threshn gt is the threshold value and represents any indicators for that zone at the time of the alarm The possible indicators are Norm Curr and Prev For example Norm lt 75 00 gt Curr lt 95 00 gt Prev This string indicates a warning 2 situation where the parameter is currently in the second zone which is not the normal zone and was previously in a critical 3 state TQ 01011 18 J 5 2200 SNMP Trap Alarms J 5 J 6 DateTime This is set to an ASCII representation of the date and time that the alarm was raised For example 01 13 2012 02 45 00 Textl This is set to the contents of User Text 1 if that field did not contain this SNMPTrap command Text2 This is set to the contents of User Text 2 if that field did not contain this SNMPtTrap command Text3 This is se
50. LC Activity Reports logical channel information for each logical channel selected DSK SUM RPT Disk I O Summary Reports disk summary statistics XPC SUM RPT XPC Summary Contains overall XPC performance information XITP SUM RPT XIIP Summary Contains overall XIIP performance information MEM WS RPT Memory Working Sets Gives information about memory working sets TQ 01011 18 Predefined Report Descriptions B 4 File Name Description File Name Description PAGE ACT RPT Paging Activity Provides a summary of paging activity MEM UTIL RPT Memory Utilization Reports memory utilization information IOTRACE Reports The IOTRACE reports described in this subsection provide information about the workloads defined for the system The IOTRACE report definition files are found in the iotrace subdirectory Note File Name Description File Name Description File Name Description File Name Description TQ 01011 18 These reports are not available for TeamQuest Online IOREQS RPT IO Requests Gives statistics for each workload on the number of I O requests I O requests that were reads and I O requests that were cached WRDXFER RPT Words Transferred Gives statistics for each workload on the number of words per input request and the number of words per output request SVCTIME RPT Service Time Gives statistics for each workload on th
51. NE A NE SOS SENG NG To perform a maintenance task enter its associated number at the prompt To return to the Administration Menu enter 99 When you select the various maintenance tasks for the Alarms and Thresholds Menu the tqzip program prompts you to supply values or other information Default values are displayed for each prompt in angle brackets lt gt To use a default press the transmit key without entering a value TQ 01011 18 7 5 Customizing Alarms and Thresholds 7 4 7 6 Creating Threshold Definitions To perform threshold checking on a given parameter or derived statistic in the performance database you must create a threshold definition for the parameter or derived statistic To create a threshold definition select Define Thresholds from the Alarms and Threshold Menu see 7 3 by entering 1 at the prompt The tqzip program displays prompts as shown in the following example gt Enter your selection gt 1 gt gt Create threshold Update threshold or Exit c u lt e gt gt c gt gt System Name gt 1 DERIVED gt 2 CLO1 gt 3 CLA gt gt Enter number name lt blank gt for more or E to exit gt 2 gt Statistic Name gt 199 Output message requeues gt 200 Output messages gt 201 Outputs queued to CMS gt 202 Page 1 0 rate gt 203 Page que gt 204 Pass off input message gt 205 Pass off messages gt 206 Percent Busy gt 207 Percent Load Fail gt 208 Pe
52. Name Description UNLKTOT RPT UDS Unlock Counts Provides total counts on the various UDS RLP unlock functions These parameters are collected for XPCs when the UDSTRP SIP counter exists UNLKREQ RPT UDS Unlock Reqs Sec Provides total counts per second for the various UDS RLP unlock functions These parameters are collected for XPCs when the UDSTRP SIP counter exists MISCTOT RPT UDS Misc Total Provides total counts on the various UDS RLP function requests other than lock and unlock These parameters are collected for XPCs when the UDSTRP SIP counter exists MISCREQ RPT UDS Misc Regqs Sec Provides total counts per second for the various UDS RLP function request other than lock and unlock These parameters are collected for XPCs when the UDSTRP SIP counter exists CD LCRET RPT Cache LC RET Gives request existence times for each of the selected logical channels broken down by cache action TAP UNIT RPT Activity by Tape Unit Provides detailed I O information for each selected tape unit TAP TOPN RPT Most Active Tape Units Provides detailed I O information for the top n tape units CD DKRET RPT Cache Disk RET Provides overall request existence time information for cache disks broken down by cache action TQ 01011 18 Predefined Report Descriptions File Name Description File Name Description File Name Description File Name Description File Name
53. PRODUCTION o cca __ tee on New mot Ewo Delte EN vw Cancel Derived Statistics PRODUCTION Help HYSTAT Log workload Total Open ERCC SUPs sec 10 SUPs sec e IP SUPs sec Throughput txns sec Population active txns Import Export Termination sec txn Lock Wait sec txn FCSS Reads sec FCSS Writes sec Delete Statistic Key In the Derived Statistics list box PRODUCTION is the database name HVSTAT Log is the category group key Workload is the category key Total is the subcategory key and CMS Queue sec txn is the statistic key Because CMS Queue sec txn is selected the status bar shows that it is the statistic key TQ 01011 18 8 7 Customizing Derived Statistics 8 3 3 Creating a New Derived Statistic Definition To create a new derived statistic definition click the New button in the Derived Statistics Maintenance dialog box see 8 3 1 To create a derived statistic definition do the following 1 In the Derived Statistic Definition dialog box enter a description of the derived statistic in the Statistic Info edit box 2 Click the Define button Using the statistic hierarchy access the desired statistic in the database and provide an alias name to be used in the definition of the derived statistic Do this for each statistic you are using 3 Use the Formula edit box to develop the derived statistic Use the alias names in the mathematical expressio
54. SYS LIB TQ BASELINE tqusrprb gt a 10 minute c Database b 0verflow end gt FIN The tqusrprb program will create the following set of parameters system User Defined Database Overflow pages PRIMUS D01 system User Defined Database Overflow Total words PRIMUS DO01 system User Defined Database Overflow Unused PRIMUS DO01 system User Defined Database Overflow Total Records PRIMUS DO1 9 44 TQ 01011 18 9 11 1 1 TQ 01011 Customizing Data Collection Probes The tqusrprb program determines the parameter hierarchy based on the options of the tqusrprb processor call and the statistic names from the TQARG file Here the system is the host system id Category group defaults to User Defined Category is taken from the c option and is Database Subcategory is taken from the b option and is Overflow The statistic names are taken from the first line of the TQARG file Each time this runstream is executed tqusrprb will store the parameter values into the 10 minute aggregation set at the current system time The DATE and TIME fields provide the timestamps If these fields are not provided tqusrprb will use current date and time as the timestamp Data Storage With all TeamQuest Baseline probes that store data into aggregation sets a data sample stored at time x represents the data for the interval starting at time x rather than the interval ending at time x The DATE and TIME given as input to tqusrprb are co
55. Short_Int_Type U_Short_Int_Type Float_Type Double_Type String_Type Byte_Array_Type 9 52 TQ 01011 18 Customizing Data Collection Probes data subtype Can be one of the following optional See 9 11 2 5 for a description of these data subtypes No_Subtype default Boolean_Subtype Time_Subtype Date_Subtype Time_Date_Subtype Elapsed_Time_Subtype Binary_Subtype Octal_Subtype Hexadecimal_Subtype Percentage_Subtype Max_Size max size Specifies the maximum size of the field It is only valid for String_Type and Byte_Array_Type If not specified zero which is unlimited is used Max_Size is the keyword for the item max size is the maximum size in bytes Time_Consolidation Indicates the field must match when consolidating across time If Time_Consolidation is not present the field does not have to match TQ 01011 18 9 53 Customizing Data Collection Probes Seq_Cons_Method cons method field name Specifies a sequential consolidation method for the field If Seq_Cons_Method is not present it is assumed that the field cannot be consolidated Seq_Cons_Method is normally used when consolidating records over time Seq_Cons_Method cons method field name is the keyword for the item can be one of the following See 9 11 2 5 for a description of these consolidation methods Id_Cons Default Min_Cons Max_Cons Sum_Cons Weight_Cons First_Cons Last_Cons is valid only for Weight_Cons types and is the field to use
56. TeamQuest software products for ClearPath OS 2200 systems TeamQuest View User Guide TQ 01401 This guide describes how to use the TeamQuest View software TeamQuest View is the user interface to TeamQuest Baseline and TeamQuest Online You use TeamQuest View to view reports from data collected by the TeamQuest Performance Software applications TQ 01011 18 vii About This Guide Unisys Product Documentation If you are using Unisys manuals you can find related product information in the following documents Business Information Server for ClearPath OS 2200 Administration Guide 7831 9415 This guide provides information system configuration database management user and run registration and configuring communications links with other Business Information Server systems ClearPath OS 2200 Cipher Application Programming Interface API Programming Reference Manual 3826 6110 This manual provides specific information on the Cipher API product used for FIPS 140 2 compliant encryption of data If installed the user id and password supplied in the TQTCP configuration element can be encrypted for security reasons for TeamQuest Baseline and TeamQuest Online ClearPath OS 2200 Exec System Software Administration Reference Manual 7831 0323 This manual provides specific information for setting up and administering the OS 2200 system This information includes instructions for booting initializing monitoring reconfiguring a
57. TeamQuest Online This separately packaged product contains additional probes to gather and store data from other system monitors It provides probes for the Message Control Bank MCB the Universal Data System UDS and mass storage See the OS 2200 TeamQuest Probes End Use Reference Manual for more information The TeamQuest PMLog software provides an additional probe for the TeamQuest Baseline software It gathers ClearPath OS 2200 Performance Monitor log data for TeamQuest Baseline See the OS 2200 TeamQuest PMLog End Use Reference Manual for more information Performance Database The performance database holds all data used to generate history reports and to monitor the system in real time Data is stored into various aggregation sets in the database An aggregation set defines the level of granularity of the data and the length of time data is stored The performance database which is maintained automatically on a daily basis is created during the initial product installation and set up process Additional databases can be created TQ 01011 18 Introduction 1 6 1 3 Host to TeamQuest View Interface 1 6 2 1 7 The host to TeamQuest View interface is called tqdata You use the TeamQuest View user interface software to connect to the host system using a socket interface A run is automatically started on the host system which executes tqdata and handles all communications between the host and TeamQuest View TeamQuest
58. The tqzip program lists the tables defined in the database and asks you for the name of the table for which you want to delete records gt Enter your selection lt 1 gt gt 4 gt gt 1 2200 Runs gt 2 Alarm Alarm Log gt 3 2200 Transactions gt Enter Table Number lt enter by name gt gt gt Class gt 2200 gt Name gt Runs The list of tables depends on the particular types of data being stored in tables Enter the number of the table whose records you wish to delete If you do not enter anything tqzip asks you for the complete table name including the table class and table name as shown in this example After you have selected a table tqzip displays the following Data Deletion Menu gt Data Deletion Menu 2200 Runs gt gt 1 Delete Records Before Time gt 2 Delete Records for System gt 3 Delete All Records gt gt 99 Exit Data Deletion Menu gt gt Enter your selection lt l gt Using the selections of this menu you can delete records based on time based on the system or all records in the table Note Data deletion takes place immediately There is no opportunity to stop the deletion of records from a table Since data deletion takes place immediately it is recommended that you back up your database before you delete any records from a table 11 16 TQ 01011 18 Customizing Tables 11 7 1 Deleting Records Based on Time The tqzip program allows you to delet
59. View Software The TeamQuest View software provides the user interface to the TeamQuest Baseline and TeamQuest Online host software TeamQuest View lets you access the host system retrieve data and create reports Using TeamQuest View you can generate reports showing historical data and view system performance data by using the real time monitor Each of the probes is released with a set of predefined reports The SIP reports give a general overview of system performance on your system The TPM reports show high volume TIP HVTIP usage by workload The IOTRACE probe reports I O information by workload The MAPPER log reports show MAPPER accounting log data by workload The system log reports show the usage of system resources by workload Product Compatibility When using TeamQuest Probes or TeamQuest PMLog with TeamQuest Baseline or TeamQuest Online you must ensure that the base levels of the products are the same The result of mixing base product levels could be database corruption For example you can run a level 7R3 product level with a level 7R3A product but it is incorrect to run a level 7R2 product with a level 7R3 or a level 7R3A product TQ 01011 18 1 11 Section 2 TeamQuest Baseline and TeamQuest Online Concepts 2 1 2 2 2 2 1 Overview The TeamQuest Baseline consists of the components that collect store manage and administer the performance data This section describes the major concepts involved Perform
60. You can then reload the workload definitions by using the l and w options on the tqzip processor call see 5 4 2 To unload workload definitions with tqzip use the following processor call format tqzip L d databasename u w workload file where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 u specifies that an unload operation is to be performed by tqzip The type of unload operation is workloads as indicated by the w option w unloads the workload definitions from the database to the file specified by the workload file parameter Example The following shows an example of unloading workload definitions to the file wfile gt tqzip u w wfile gt TQZIP 7R4G 011312 1104 55 Fri Jan 13 15 13 19 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt gt Using database TQDATA PRODUCTION gt gt Workload definitions unloaded gt gt End TQZIP TQ 01011 18 5 19 Customizing Workloads Workload Definition File Format The workload set and workload definitions appear in the following format in the output file created by tqzip If you choose to edit this file or update the definitions you should maintain this file format BEGIN WORKLOADS unloaded on Fri Jan 14 10 54 13 2012 A sys
61. a database enter the following command tqdbu i database name where database name is the name of the new database Database names can be any alphanumeric name up to 12 characters in length The names must be unique within the first four characters After creating a database you may need to customize your data collection runstreams and tq ini element to make use of the new database see Section 9 Deleting Databases TeamQuest Baseline and TeamQuest Online enable you to delete a database when you no longer need it Use the following call statement to delete a database tqdbu r database name where database name is the name of the database to be deleted After deleting a database you must update your data collection runstreams and tq ini element to remove references to the deleted database see Section 9 TQ 01011 18 10 3 Customizing Performance Databases 10 6 10 7 10 8 10 4 Listing Existing Databases You can use the tqdbu utility to obtain a list of the databases that currently exist on your system Enter the following command tqdbu 1 tqdbu then displays a list of the performance databases that exist on your system Estimating Disk Space Requirements If you are using TeamQuest Baseline the amount of disk space occupied by a performance database depends on a number of different factors mainly the number of devices available on the system The files that grow the most are those files containing t
62. a workload the data is assigned to only that workload The remaining workloads are not evaluated If the record does not belong to any defined workload it is assigned to a default workload named OTHER The sum of all workloads including OTHER represents all the work performed on the system You define workloads using the workload definition language see 5 2 2 To create workloads you enter the definitions by using TeamQuest Performance Software Administration provided by TeamQuest Baseline or by updating a file unloaded by the tqzip program see 5 4 1 TQ 01011 18 5 1 Customizing Workloads 5 2 1 5 2 2 Workload Qualifiers Each probe allows you to identify specific qualifiers for each workload set Depending on the source data type you can qualify the workload sets according to system name environment name or DCP name The system activity monitor TPM and system log probe workload sets can be qualified by system name If no qualifying value is defined for the workload set the workload set is assumed to apply to all systems whose data is analyzed by the system activity monitor or system log probes The MAPPER log probe workload sets can be qualified by system name and MAPPER environment name If no system name is specified the workload set is assumed to apply to all systems If no MAPPER environment name is specified the workload set definition is assumed to apply to all MAPPER environments whose data is analyzed by th
63. activities are deactivated on any available queue Number of available queues in the system Queue Number of banks in the system TQ 01011 18 OS 2200 Parameters A 3 TPM Log Parameters This subsection contains a list of the parameters collected by the TeamQuest Baseline TPM and RTPM log probes The following statistics are collected for each workload defined on the system Note These parameters are not collected by TeamQuest Online Parameter Hierarchy Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs T O SUPs Lock Wait Time Lock Held Time Processing Time Response Time Termination Time Initial Loads Initial Load Time Stuck Copies Total Stuck RTPS Deact Copies Total RTPS Deact Active Copies Total Active Txns Scheduled Txns Completed Passoff Msgs Output Messages TQ 01011 18 HVSTAT Log Workload Total Total standard units of processing SUP Total IP SUPs Total ER and control card SUPs Total I O SUPs Total time spent waiting on locks Total time spent holding file control superstructure FCSS locks Total amount of time spent processing Total amount of time spent to complete a transaction The average commit duration time The number of transaction initial loads and block transfer loads The total amount of time between program initial load and initiation for TIP initial loads and block transfers Total number of stuck transactions with
64. aggregation set 6 14 derived statistic 8 16 tables 11 4 workload definition 5 16 workload set definition 5 11 E encrypted connections TQENC program I 10 user id and password I 10 using I 1 encryption in TQTCP I 10 TQ 01011 18 errors for TQTCP I 9 exporting data using tqrep E 12 derived statistic definitions 8 15 8 18 threshold definitions 7 21 workload definitions 5 17 5 19 extracting statistical data F 1 G general section report definition file C 3 H hierarchy derived statistic 8 6 host software data collection by 1 8 description 1 8 probe 1 8 host based reporting aggregation set data D 1 table data E 1 importing derived statistic definition 8 15 workload definition 5 17 installing tqusrprb probe 9 47 IOTRACE data collection description 9 25 parameters A 44 reports B 13 starting 9 25 T O activity statistics A 24 T O reports B 2 K key names description 2 2 viewing 8 14 Index 3 Index L listing aggregation set 6 20 collection periods for tables 11 6 databases 10 4 parameters 10 4 retention specifications for tables 11 11 loading derived statistic definitions 8 21 threshold definitions 7 24 workload definitions 5 22 maintaining alarms 7 1 tables 11 1 11 3 thresholds 7 1 management console configurations Computer Associates configuration J 9 HP OpenView configuration J 7 overview J 6 Tivoli configuration J 7 MAPPER log data collection des
65. and RSI would automatically generate the site ids for each connection Unique User ids For Each TQTCP Run The user id and account are the only statements that are required The password used to start the TQTCP run is solicited by TQ View or TQ Administration and should not be contained in the TQTCP configuration element All user ids that are allowed to run either TQ View or TQ Administration must exist in the TQTCP configuration element interface interface type This tells TQTCP the communications interface to use when listening for incoming connections The interface type can be either CMS1100 or CPCOMM If this statement is not supplied the default is CPCOMM port port number This tells TQTCP the port number to listen on for incoming connections If this statement is not supplied the default port number is 2700 siteid xxxxx Optional list of site ids to use for RSI runs If not present site ids will automatically be generated by RSI Site ids are specified one per line Up to 50 concurrent RSI runs can be executing within TQTCP Each socket connection initiates an RSI run If using specific site ids TQTCP will use the first 50 site ids specified here If there are more the additional ones are ignored If there are fewer the number of site ids specified is used as the maximum number of concurrent RSI runs xqtfile qualifier filename This file is where the tqdata executable resides This file is used in the RSI run to execu
66. and Workload Definitions 5 3 9 Deleting a Workload ou eeeeeeceeceeeteeeeeeeeneeeeeeeeeaes as Wass Ws Vics A ankARwWwW 4 2 4 3 4 5 4 6 4 6 4 6 4 7 4 8 ana aan nn ROMNM aan NNO on O1 It tt JT tag gt ONNDWN O0O OO ana nan on ee h h ee ee TQ 01011 18 Contents Section 6 Section 7 TQ 01011 18 5 4 Maintaining Workloads with tqZip seese 5 4 1 Unloading Workload Definitions oo 5 4 2 Loading Workload Definitions n se Customizing Aggregation Sets 6 1 laiiKelo OTeiilo a MAE a a E A eee A 6 2 Concepts aiahsiheniatihesdieviah a a AE cies 6 2 1 Aggregation Set Structure o ecceeeecsseeeeeeeenees 6 2 2 Moving Window Concept for Data Storage _ 6 2 3 Aggregation Set Consolidation cceeeeees 6 2 4 Feeding Aggregation Sets 6 2 5 Compact and Expanded Aggregation Set Format 6 3 Maintaining Aggregation Sets with TeamQuest Performance Software Administration 06 6 3 1 Performing Aggregation Set Administration Tasks 6 3 2 Creating an Aggregation Set sses 6 3 3 Displaying or Modifying Aggregation Set Definitions 6 3 4 Deleting an Aggregation Set esee 6 4 Maintaining Aggregation Sets with tqasa_ eeeee 6 4 1 Creating an Aggregation Set n se 6 4 2 Removing an Aggregation Set eceeeeeeeee 6 4 3 Listing Aggregation Sets n 6 4 4 Unloading Aggregation Set Definitions 6 4 5 Loading Aggreg
67. annotation is printed Background Indicates the color of the best fit annotation box that appears directly behind the best fit annotation text If a background color is specified the best fit annotation box covers any graph lines and other text that may appear behind this area If the Background field is not specified TeamQuest View allows any graph lines or other text that may be behind the title to show through Border Indicates the color of the border line that appears for the background area of the best fit annotation text If the Border field is not specified TeamQuest View does not display a border line around the background area of the best fit annotation text The default values for the text attributes are Arial Normal 8 and Black TQ 01011 18 Report Definition File Note Text Top Left Bottom Right Vertical Horizontal Font Style Size Foreground Background Border The specifications for annotation that appears on the report The text of the annotation appears in a rectangle defined by relative coordinates with a range of 0 through 10 000 Text Can be up to 80 characters long If it contains spaces it must be enclosed in quotation marks Top Specifies the relative location for upper coordinate of the top left hand corner of the annotation box Left Specifies the relative location for the left coordinate of the top left hand corner of the annotation box Bottom Specifies the relative lo
68. any number of the character O and ending with T This matches ROT ROOT ROOOT and so on e regexp is a regular expression that matches the regular expression regexp This rule can be used to build complex regular expressions For example AB matches a null string AB ABAB ABABAB and so on 5 6 TQ 01011 18 5 2 2 4 5 2 2 5 Customizing Workloads Using Boolean Expressions in Definitions The Boolean operators OR AND and NOT can be used to combine numerical conditions and string conditions to form a definition Parentheses can be used to override the default precedence The following example shows the use of Boolean operators PID 111 OR Account 10100 AND Run_ID TQ In this example the numerical condition PID 111 must be true or both string conditions Account 10100 and Run_ID TQ must be true The operators use the following order of precedence Operator Order of Precedence 0 left to right lt gt lt gt lt gt left to right p left to right NOT right to left amp amp AND left to right OR left to right The operators and are equivalent The operators lt gt and are equivalent The operators AND and amp amp are equivalent The operators OR and are equivalent The operators NOT and are equivalent Definition Example The following example shows a definition using characteristics for run identifiers account and PID Run_ID VENCTL OR
69. at the top of the dialog box The first item in the Workload Info list box is a source data type Source data types are defined when a probe is registered with the database at installation time They automatically appear in the Workload Info list box Source data types cannot be modified and therefore the Open and Delete buttons do not apply to the source data types The items listed under the source data type are the workload sets and the workloads Workload sets are listed at the first indentation under the source data type and workloads are indented under the workload set to which they belong TQ 01011 18 5 9 Customizing Workloads 5 3 2 5 10 The status bar at the bottom of this dialog box displays information about the selected item in the Workload Info list box It tells you the type of item source data type workload set or workload that is selected from the list You can use the arrows to the left of the Workload Info list box to rearrange the workloads within the workload set for which they are defined Select the workload and click the up or down arrow to move the workload up or down the list See 5 2 for a description of how the order of the workloads affects the processing of data Creating a New Workload Set Definition To create a new workload set definition select the source data type from the Workload Info list box in the Workload Maintenance dialog box Notice that the term workload set appears under the Ne
70. box The Workload Definition dialog box is displayed 4 Workload Definition Session cl01 Database PRODUCTION Source Data Type syslog Workload Set Name HR wis Cancel Workload Name HATSO Help Workload Characterization Characteristics Transaction HR700 HA571 Trans_Code lt Operators Logical Operators You can use this dialog box to modify the workload name and workload characterization For a description of the fields in this dialog box see 5 3 5 When you click OK your changes are accepted and you return to the Workload Maintenance dialog box see 5 3 1 Your changes are not sent to the host software until you click OK in the Workload Maintenance dialog box 5 16 TQ 01011 18 5 3 7 5 3 8 Customizing Workloads Importing Workload Sets and Workload Definitions To import all workload set and workload definitions click the Import button on the Workload Maintenance dialog box see 5 3 1 When the Open File dialog box is displayed select the file that contains the workload set and workload definitions that you want to import When the workload set and workload definitions are imported all existing workload sets and workload definitions are updated to match the imported definition If you deleted a workload definition from the exported file and then imported the remaining workload definitions the deleted workload definitions will be re ordered to the last position Use the Delete bu
71. correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The tqzip program displays the list of threshold definitions currently defined in the performance database Select the definition you want to view You can enter either the number of the parameter or the full name at the prompt Derived statistic threshold names begin with DERIVED B The selected threshold definition is displayed c Enter y at the prompt to view each qualifier list for this derived statistic threshold The qualifier lists for system names and resource names are shown in the example D If you enter y at the final prompt the tqzip program prompts you for another parameter name to view If you enter n at the prompt the tqzip program returns to the Alarms and Thresholds Menu see 7 3 7 11 Deleting Thresholds The tqzip program allows you to delete an existing threshold definition Select Delete Threshold on the Alarms and Thresholds Menu see 7 3 by entering 3 at the prompt The tqzip program then prompts you with a list of parameter and derived statistic names that currently have threshold definitions in the database as shown in the following example gt Enter your selection gt 3 gt gt Current Threshold Definitions gt 1 DERIVED Derived Derived Category Derived Subcategory UserCpuUtil A gt 2 DERIVED Derived Derived Category Derived Subcategory ResplTimeGoal gt 3 CLO1
72. d option is not specified the Database setting in the tq ini element is used see 3 1 Limit the action specified by the c or l option to the specified run id Do not use the r and g options together in one execution of the tqunlk program Limit the action specified by the c or l option to the generated run id of this run Do not use the r and g options together in one execution of the tqunlk program Use with the l option to obtain a list of any locks currently being held for all database qualifiers database names and run ids Backup and Recovery You should save and restore the contents of the individual files that make up the performance database as a group If for some reason you need to restore one of the database files from a backup you should restore all of its companion files from the same backup Caution Back up the performance database periodically You might have to restore the database in the case where a TeamQuest Baseline probe or the tqagg program is abnormally terminated and leaves the database files in an inconsistent state TQ 01011 18 Section 11 Customizing Tables 11 1 Introduction The tqzip program is used to maintain the tables in your performance database If you are happy with the default values used no additional administration is required Defaults are set by the probes when the tables are initially created However if you want to customize your data collection you can p
73. database you must click OK in the Aggregation Set Maintenance dialog box You can cancel the deletion of an aggregation set either by clicking the No button in the alert box or by clicking Cancel in the Aggregation Set Maintenance dialog box see 6 3 1 Note When you remove an aggregation set the aggregation set file is deleted from the database Maintaining Aggregation Sets with tqasa The tqasa program provides an optional method for performing the following tasks associated with maintaining aggregation sets e Create aggregation sets see 6 4 1 e Remove an aggregation set that is no longer needed from the database see 6 4 2 e Display information about the aggregation sets defined for a database see 6 4 3 e Unload aggregation set definitions in a database see 6 4 4 e Load aggregation set definitions in a database see 6 4 5 e Convert an existing expanded aggregation set into compact format see 6 4 6 After you create remove or change an aggregation set you must customize your data collection to take advantage of any aggregation changes see 6 4 7 Note The capability to maintain multiple aggregation sets is available only with TeamQuest Baseline TeamQuest Online allows only one aggregation set in the database After you create remove or change an aggregation set you may need to update the TeamQuest Baseline probe runstreams in SYS LIB RUN or the tq ini element in SYS LIB TQ BASELINE If you want the probes t
74. database using the file specified by the parm file parameter See 10 9 The parm file used with the p option must be in the same format as the file produced by the o option All key values must be specified in each parameter reference specifies that a load operation is to be performed This option must be used with one of the following options c r s t or w Information contained in the file specified by the c r s t or w options is loaded into the database specifies that an unload operation is to be performed This option must be used with one of the following options c r s t or w Information contained in the file specified by the c r s t or w options is unloaded from the database tqzip Processor Call Format e is used in combination with the l option when loading workload reduction threshold or derived statistic definitions If you specify the e option the new definitions are loaded without first deleting the existing definitions in the database If you do not specify the e option all existing definitions are deleted from your database before the new ones are loaded C specifies the file that tqzip should use for a table collection retention load or unload operation The tablespec file parameter is the name of the file containing the data collection and retention specifications for your database The c option must be used with the l or u option See 11 10 and 11 11 r specifies th
75. determine whether to terminate or continue with data collection TQ 01011 18 Customizing Data Collection Probes You can put up to ten tables in a table file It is a good idea to put similar tables in the same file The TeamQuest Baseline table manager handles many operations on a file level and it is not a good idea to have a very large number of files open at one time This example assumes that you will put all of the user created tables in a file called usertabs Only one table can be created with a single tqtblprb i execution Table Storage The syntax for executing tqtblprb to store data into a table is filename tqtblprb f file d databasename s system t class name end where f specifies the name of a file that contains the data to be stored If the f option is specified the data is read from the file If the f option is not specified it is read from the runstream d specifies the database into which data is stored The databasename parameter is the name of an existing database on your system If the databasename is not specified the database in the tq ini file is used S specifies the system to use for data storage If the s option is present the specified system is substituted for any occurrence of system If the s option is not present the system name found in the program s initial register set is used t specifies the table class and table name for data collection and storage This field is
76. edit boxes A list of all possible values is displayed If the value you want the qualifier to take on is not provided in the list you can type in your desired value The View button allows you to look at long key names that may have been truncated in the list boxes see 8 3 5 Deleting an Alias To delete an alias from the Alias List edit box select the alias from the Alias list box of the Derived Statistic Definition dialog box and click the Delete button The alias is removed from the list box The alias name is not removed from the Formula edit box in the Derived Statistic Definition dialog box You must remove those references yourself Saving a Derived Statistic Definition To save a newly defined derived statistic the Save Statistic dialog box is displayed when you click the Save As button from the Derived Statistic Definition dialog box or when you click OK for a new derived statistic definition This dialog box allows you to specify where to place the new derived statistic in the statistic hierarchy Save Statistic Statistic Statistic Hierarchy PRODUCTION Derived Processor I0 Activity Memory System Log HYSTAT Log Mapper Log System UDS MassStor by System MassStor by Equip MassStor by Disk MCB TQ 01011 18 8 13 Customizing Derived Statistics Save Statistic Dialog Box The Save Statistic dialog box contains the following fields Statistic enter the name for your derived s
77. example you might want a single point for each shift of a work day The shifts run from 7 00 to 15 00 15 00 to 23 00 and 23 00 to 7 00 the next day In this case you would want the synchronization time to be 7 00 though it could be specified as 15 00 or 23 00 with the same results The primary retention contains data at whatever interval it is collected You can only modify the retention period and not the interval of the retained data If the data stored in the table is not intervalized such as the Alarm Log no retentions other than the primary retention can be defined To begin the administration of the retention specification for a table select Retention Periods from the Tables Menu see 11 3 by entering 3 at the prompt The tqzip program lists the tables defined in the database and asks you for the name of the table whose retention specification you want to administer gt Enter your selection lt 1 gt gt 3 gt gt 1 2200 Runs gt 2 Alarm Alarm Log gt 3 2200 Transactions gt Enter Table Number lt enter by name gt gt gt Class gt 2200 gt Name gt Runs The list of tables depends on the particular types of data being stored in tables Enter the number of the table whose collections you wish to administer If you do not enter anything tqzip asks you for the complete table name including the table class and table name as shown in this example 11 10 TQ 01011 18 Customizing Tables After y
78. executable back into that file TQ 01011 18 7 31 Customizing Alarms and Thresholds 7 20 Single Point Operations SPO 7 32 Considerations Note This subsection applies only if you use the Unisys Single Point Operations SPO products The SPO products monitor and control multiple computer systems from one operations console The message automation and alarm handling functionality of SMART or the Single Point Autoaction Message System SP AMS and the alarm capability of TeamQuest Baseline can be integrated by following the guidelines listed in this subsection The two key components of integrating TeamQuest Baseline and SPO are the TeamQuest Baseline threshold definition and the associated SMART SP AMS pattern that is used to process the console alarm message generated when an alarm is raised or cleared by TeamQuest Baseline SMART or SP AMS transforms the TeamQuest Baseline message into an event report that tells SPO and Status the necessary information to create displays and carry out automated alarm actions In addition SMART or SP AMS may directly take automated actions based on the message The alarm identifier and the user text lines specified when defining thresholds with tqzip see lines N and 0 in 7 4 are used to build alarm messages that are sent to the system console by TeamQuest Baseline Certain conventions should be followed when defining thresholds using tqzip to ensure easy pattern matching and data extraction us
79. field and thus the record to be rejected The value must be a valid short integer U_Short_Int_Type Subtype is not checked for this type Allowable characters are numeric digits and the plus sign if it precedes the numeric digits Any other character will cause the field and thus the record to be rejected The value must be a valid unsigned short integer Long_Int_Type Subtype is not checked for this type Allowable characters are numeric digits the plus sign and hyphen if they precede the numeric digits Any other character will cause the field and thus the record to be rejected The value must be a valid long integer U_Long_Int_Type Subtype is checked for the type and determines the data format rules as follow Time_Subtype Time_Subtype is stored internally as seconds past midnight and the allowable input formats are hh mm ss hh mm SS Date_Subtype Date_Subtype is stored internally in time_t format and the allowable input format is madd yylyy Time_Date_Subtype Time_Date subtype is stored internally in time_t format and the allowable input formats are mm dd yylyy hh mm ss time Elapsed_Time_Subtype Elapsed_Time_Subtype is stored internally as seconds and the allowable input formats are hh mm ss hh mm SS TQ 01011 18 TQ 01011 18 All Other Subtypes Float_Type Double_Type String_Type Byte_Array_Type Customizing Data Collection Probes Allowable characters are numeric d
80. following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The tqzip program prompts you for the file name for your alarm messages file B The tqzip program also prompts you for a starting date and time as well as an ending date and time You can limit the number of messages that are output to the alarm messages file by specifying a starting and ending date and a starting and ending time If you do not enter a starting date or time tqzip uses the first date and time found in the alarm file If you do not enter an ending time all messages after the starting time are output c The tqzip program generates the list of alarm messages and places them in the alarm messages file you specified as the output file name See 7 18 for a description of the format of alarm messages TQ 01011 18 7 27 Customizing Alarms and Thresholds Alarm Messages File Format The following shows how the alarm messages appear in the output file gt All available alarm messages follow gt gt Fri Jan 13 08 25 33 2012 gt TQ CRITICAL lt 12 51 gt CLO1 Demand Response Time gt Call John at extension 2332 gt gt Mon Dec 19 15 29 02 2011 gt TQ CRITICAL lt 43212 00 gt CLO1 Fixed Tracks Available gt gt End of alarms 7 17 Using tqalm The TeamQuest Baseline probes perform threshold checking for all parameters that have a threshold definition in
81. handling will ensure that any locks held by the probe are cleared If the contingency handler does not receive control on abnormal termination locks may not be cleared The tqunlk program is provided for clearing any locks left by abnormally terminated probes This processor is executed in each of the probe runstreams immediately following the probe processor call statement This execution will ensure that all locks created by the generated run id of the probe run are cleared Normally you should never need to execute tqunlk to clear locks If you need to use tqunlk be careful to ensure that you are not clearing a lock that is required by a probe that is currently executing on your system TQ 01011 18 10 7 Customizing Performance Databases 10 11 The tqunlk processor call has the following format where C L q qualifier d databasename r runid Clear any locks currently being held Either the c option or the l option must be specified However do not use the c and l options together in one execution of the tqunlk program List any locks currently being held Either the c option or the l option must specified However do not use the c and l options together in one execution of the tqunlk program Use the specified database qualifier The default qualifier is TQDATA Use the specified database name The databasename parameter is the name of a database that has been created on the host system If the
82. has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 a indicates the action to be taken with the threshold checking setting action can be one of the following ON turns threshold checking on OFF turns threshold checking off LIST displays the current threshold checking setting 7 26 TQ 01011 18 Customizing Alarms and Thresholds 7 16 Generating Alarm Lists You can look at the alarm messages that have been generated by the TeamQuest Baseline probes by using the Generate Alarm List selection on the Alarms and Thresholds Menu see 7 3 The tqzip program generates a list of the alarm messages and places them in a specified file In order to generate alarm lists through tqzip the tqalm program must be executing on your system See 7 17 for information on how to start tqalm To view the alarm messages with tqzip select Generate Alarm List from the Alarms and Thresholds Menu see 7 3 by entering 6 at the prompt The following example shows the prompts displayed when you select Generate Alarm List from the Alarms and Thresholds Menu Enter your selection 6 Enter the output filename A alrmfile Enter start date amp time yy mm dd hh mm ss B ter end date amp time yy mm dd hh mm ss gt gt gt gt gt gt gt gt gt gt gt outputting alarm messages to alrmfile c gt gt 2 alarm messages were selected The
83. input from the TQARG file It then reads the data values as input from the PRBDATA file Using the tqusrprb probe processor call options and the input the tqusrprb probe creates parameters and stores values corresponding to those parameters in the performance database By default the tqusrprb probe reads the tq ini file to determine the database and aggregation sets in which to store data Processor Call Format The processor call format of tqusrprb is sys lib tq baseline tqusrprb p d databasename s system g catgroup c category b subcat L a aggsetl1 h END where p specifies that the timestamps supplied in the PRBDATA file represent point in time timestamps and data is stored within the corresponding aggset interval Without the p option data is stored one interval earlier to match data collected by other probes such as tqbsp d databasename specifies the database into which data is stored The databasename parameter is the name of an existing database on your system If the d option is not specified the Database setting in the tq ini element is used see 3 1 s system specifies the system key name in the parameter hierarchy of the statistic being stored If the system is not specified the name of the current system as returned in the program s initial register set is used g catgroup specifies the category group key name in the parameter hierarchy of the statistic being stored
84. is not specified 000000 is assumed The special identifier amp NDATE can be used to specify the current date amp NTIME can be used to specify the current time amp NTDATE can be used to specify the current date and time EndTime MMDDYYYY HHMMSS The ending date and time for the report This specification can be overridden by the e option on the tqrep command line If the time HHMMSS is not specified 000000 is assumed The special identifier amp NDATE can be used to specify the current date amp NTIME can be used to specify the current time amp NTDATE can be used to specify the current date and time TimeCons nn units The time consolidation to use on reported data nn is the number of units units can be HOUR MIN or SEC This specification can be overridden by the c option on the tqrep command line If TimeCons key is not specified in the report definition file and the c option is not given on the command line no time consolidation is performed Host Based Reporting for Table Data PageLength nn The length of the printed page in lines nn is the number of lines If the PageLength key is not specified 60 lines is assumed PageWidth nn The width of the printed page in character positions nn is the number of character positions The page width can be overridden by the w option on the tqrep command line If the PageWidth key is not specified in the report definition file and the w option is not given on the com
85. items are not collected by Team Quest Online A 10 1 System Table Information The following information is stored in the performance database tables Table Field Hierarchy Class 2200 Name ODTP System File name odtpfile Collection interval Based on the table collection interval Default retention 1 day Data Item Type Description Time Double The time this information was recorded in the database for this sample Interval Double The desired sample interval in this collection period Svc_ReqQ Double The current number of service requests queued in this sample interval Config_ReqQ Double The maximum configured request queue limit Servers_Actv Double The current number of active servers doing work Servers_Wait Double The current number of servers waiting for new work Servers_Avail Double The current number of servers available and waiting for additional work for the same transaction Servers_Attch Double The current number of servers currently attached TQ 01011 18 A 61 OS 2200 Parameters A 10 2 A 62 Servers_Max Clients_Attch Clients_Max Threads_Curr Threads_Max Threads_Resv Threads_Config Double Double Double Double Double Double Double The maximum number of servers attached since the transaction management system control TMSC processor started The current number of clients currently attached The maximum number of clients attached since the transaction management s
86. material provided to the U S Government are provided and delivered subject to the commercial license rights and restrictions described in the governing license agreement All rights not expressly granted therein are reserved TeamQuest the TeamQuest logo TeamQuest Alert TeamQuest Analyzer TeamQuest Baseline TeamQuest CMIS TeamQuest Harvest TeamQuest IT Service Analyzer TeamQuest IT Service Reporter TeamQuest Manager TeamQuest Model TeamQuest On the Web TeamQuest Predictor TeamQuest Online TeamQuest Surveyor TeamQuest View and Performance Surveyor are trademarks or registered trademarks of TeamQuest Corporation in the US and or other countries Unisys and ClearPath are registered trademarks of Unisys Corporation in the US and or other countries All other trademarks and service marks are the property of their respective owners About This Guide TeamQuest Baseline and TeamQuest Online are required OS 2200 software packages that provide the services for the TeamQuest Performance Software family TeamQuest Baseline is installed on a host system and provides the components to collect store manage and administer your performance data Purpose The OS 2200 TeamQuest Baseline and TeamQuest Online User Guide describes TeamQuest Baseline and TeamQuest Online It tells you how to use TeamQuest Baseline as a tool for performing performance management and performance analysis tasks on your system and how to use TeamQuest Online as a
87. more information on workload characterization and the workload definition language Note The characteristics in the Characteristics list box are case sensitive These characteristics must appear in the Workload Characterization edit box exactly as they appear in the Characteristics list box To complete a workload you must enter a workload characterization If the field does not contain a value an error message is displayed A syntax check will be done on the workload characterization for verification Characteristics Operators and Logical Operators These list boxes aid you in creating the workload definition You can select an item from one of these boxes and then click the arrow button to transfer the item to the Workload Characterization edit box You can also double click the item to transfer it Characteristics list box contains a list of the characteristics defined for the source data type Operators list box contains a list of the operators allowed for the workload definition language Logical Operators list box contains a list of the logical operators allowed for the workload definition language TQ 01011 18 5 15 Customizing Workloads 5 3 6 Displaying or Modifying a Workload Definition To display or modify a workload definition select the workload from the Workload Info list box in the Workload Maintenance dialog box see 5 3 1 and click Open You can also double click the workload name in the Workload Info list
88. no longer needed it can be deregistered from the performance database Deregistering a file does not cause a table file to be deleted It simply removes the record of the table from TeamQuest Baseline Manual Record Deletion Table records are normally deleted from tables by the TeamQuest Baseline aggregation program based on the retention specifications for the table There may be times when you want to remove records from a table manually This can be accomplished for a given system for all records before a given point in time or for all records in the table TQ 01011 18 Customizing Tables 11 3 Opening the Table Administration Menu The tqzip program allows you to perform the various tasks associated with maintaining tables in the performance database The tqzip program displays a series of menus for you to select the tasks you want to perform When you execute the tqzip program it displays the Administration Menu that allows you to select the type of administration tasks you want to perform To perform maintenance tasks on tables you execute the tqzip program and select Tables by entering 2 from the Administration Menu gt tqzip gt TQZIP 7R4G 011312 1104 45 Fri Jan 13 09 19 55 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt gt Using database TQDATA PRODUCTION gt gt gt Administration Menu gt gt 1 Alarms and Thresholds gt 2 Tables gt gt 99 Exit
89. of the runs in a MAPPER system or locally for certain runs It is recommended that sites use the local run function logging approach if this additional information is desired 9 9 1 Data Collection Runstream The tqinit process at installation time creates runstreams tailored to the MAPPER environments on your system For every MAPPER environment to be probed a runstream named tqxmr is created Each of these runstreams sample a particular MAPPER environment and each must be started separately to collect MAPPER performance information Each runstream uses a tqxmr run identifier and executes the tqgmpr program The tqmpr program processes data from a MAPPER log and stores it in the performance database The tqmpr program processes all cycles of the MAPPER log file from the previous day During the tqinit process at installation time you are prompted for the time that the tqxmr runstream starts each day The tqxmr runstream should run each day to ensure that your performance database contains a complete set of data for each day 9 9 2 tqmpr Processor Call Statement The tqxmr runstreams contain the tqmpr processor call statement which controls the MAPPER accounting log data reduction The tqmpr processor call statement specifies data collection parameters including input parameter table name database name aggregation set name and MAPPER environment name Each tqxmr runstream contains the tqmpr processor call parameters Normally you use t
90. on IP SUPs ER and control card SUPs and IO SUPs SVCLVL RPT Service Levels by Workload Gives statistics for the selected workload and environment total TIP demand batch or deadline on total SUPs voluntary wait time resource wait time and active time TQ 01011 18 Predefined Report Descriptions B 8 File Name Description File Name Description THRUPUT RPT Throughput by Workload Gives statistics for the selected workload and environment total TIP demand batch or deadline on number of I O requests disk I O requests disk I O words transferred tape I O requests tape I O words transferred number of transactions cards the Exec read in and the number of runs opened SUMMARY RPT Information by Workload Gives statistics for the selected workload and environment total TIP demand batch or deadline on total SUPs IP SUPs ER and control card SUPs IO SUPs core block SUPs voluntary wait time resource wait time active time number of I O requests disk I O requests disk I O words transferred tape I O requests tape I O words transferred number of transactions cards the Exec read in and the number of runs opened real pages concurrently resident and real memory usage TPM Log Reports The TPM reports provide information about the workloads defined for the system The TPM log reports are found in the tpmlog subdirectory Note File Name Description File Na
91. option can be used with the t option The transactions parameter can be any number from 1 to 20 The default value is to collect information on the 20 longest running active transactions Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent lines of the processor call The parameters specified on the processor call statement must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keyword The tqsam probe has a default collection period retention period and sample interval See Section 11 for more information on maintaining table data The defaults are as follows collection period The collection period indicates a time range that the tqsam probe when running is to sample system activity By default tqsam will continuously collect data sample interval The sample interval in seconds indicates how often the tqsam probe is to sample system activity within a collection period The default sample interval is 30 seconds retention period A retention period represents how long the data is maintained in the database A retention period called Primary exists for each table and has a default retention period of 24 hours Terminating Data Col
92. or the tqasa program to delete and re create the aggregation sets in expanded format See 6 2 5 for information on compact and expanded aggregation set format e The workload probes tqtpm tqrtpm tqiot tgmpr tqsys and tqtlp only store data into the one aggregation set The name of the aggregation set must be specified on the processor call statement using the a parameter The supplied runstreams specify the 10 minute aggregation set on the processor call statements e TeamQuest Online has only one default aggregation set defined That aggregation set collects data at 1 minute intervals and maintains the data in the performance database for 32 hours Alarm Handling The TeamQuest Baseline probes perform threshold checking for all parameters and derived statistics that have a threshold definition in the performance database When an alarm condition is encountered the probe generates an alarm Alarms are handled by the tqalm program that is provided with TeamQuest Baseline By default the tqalm program sends alarm messages to the system console See Section 7 for information on how to set up and maintain your threshold definitions TQ 01011 18 3 3 TeamQuest Baseline Administration 3 3 1 3 3 2 Default Threshold Definitions A set of default thresholds is included with TeamQuest Baseline To use these default thresholds you must first load the threshold definitions into your performance database The element SYS LIB TQ BAS
93. parameters should appear on the second and subsequent lines of the processor call TQ 01011 18 Customizing Data Collection Probes 9 7 4 TPM Log File Input With the tqtpm program you can load data from a TPM log file into a performance database by specifying the date range or the specific TPM log file This enables you to load TPM log data from another system or to update the performance database with TPM log data that was collected at a time when tqtpm was not running on your system When loading data using a date range tqtpm searches the system and processes all TPM log files that contain information in the date range You must specify two additional parameters on the tqtpm processor call statement See 9 7 3 for the format of the processor call statement The format for the date range parameters is b mmddyylyy e mmddyyl yy where b specifies the month day and year of the beginning date used in the log file selection e specifies the month day and year of the ending date used in the log file selection If a 2 digit year is specified the digits 00 through 27 are interpreted as the years 2000 through 2027 and the digits 28 through 99 are interpreted as the years 1928 through 1999 To load data directly from specific TPM log files you must specify two additional parameters on the tqtpm processor call statement See 9 7 3 for the format of the processor call statement The format for these parameters is f tpmlo
94. read hit transfer time Cache disk read hit service time Cache disk read hit hardware service time Cache disk read hit percent of all requests IO Activity Cache Disk Write Hit Cache disk write hit requests per second Cache disk write hit K words per request Cache disk write hit transfer time Cache disk write hit service time Cache disk write hit hardware service time Cache disk write hit percent of all requests IO Activity Cache Disk Read Miss Cache disk read miss requests per second Cache disk read miss K words per request Cache disk read miss transfer time Cache disk read miss service time Cache disk read miss hardware service time Cache disk read miss percent of all requests TQ 01011 18 OS 2200 Parameters Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Subcategory Statistic Name Reqs Sec KWords Req TQ 01011 18 IO Activity Cache Disk Write Miss Cache disk write miss requests per second Cache disk write miss K words per request Cache disk write miss transfer time Cache disk wr
95. run started The total elapsed time for the run since the run started The interval elapsed time since the last sample for the run The time the program started Cumulative IP SUPs for this run since inception IP SUPs for this run in this interval Cumulative I O SUPs for this run since inception IO SUPs for this run in this interval Cumulative ER CC SUPs for the run since inception ER CC SUPs for this run in this interval Cumulative I O requests for this run since inception Interval I O requests for this run in this interval Total I Os per second for this run since inception Interval I Os per second for this run in this interval The number of combined occurrences The following information is stored in the performance database tables Table Field Hierarchy Class Name File name Collection interval Default retention Data Item Run_ID Transaction Elapsed TQ 01011 18 2200 Transactions samfile Based on the table collection interval 1 day at primary retention specification Type String String Double Description The generated run id of the transaction The name of the transaction program being executed The elapsed time from when the transaction program started to the current time A 59 OS 2200 Parameters A 60 Tot_Sups KSize Type Start Time Sample_Time Interval System Tot_IP Tot_IO Tot_ER CC Tot_IOreq Count Double Integer String Double Do
96. runs MSCP RUNNER CMR PCCTI and DDP By using the Run_ID MSCP Run_ID MSCP construct data for runs with identifiers of MSCP and MSCPx where x can be any single alphanumeric character is included in this workload Run_ID MSCP Run_ID CMR Run_ID PCCTI Run_ID DDP TQ 01011 18 Run_ID MSCP Run_ID CMR Run_ID PCCTI Run_ID DDP Run_ID RUNNE 9 37 Customizing Data Collection Probes 9 11 9 38 Example 2 The following example shows a workload definition using run identifiers and account numbers to identify the workload This workload contains information for all runs with identifiers that start with TQ and use account numbers in the range of 123100 123199 The workload also contains information for run TQSYS By using the Run_ID TQ construct runs with identifiers for TQ and TQx where x can be any string of alphanumeric characters are considered for inclusion in the workload By using the Account 1231 construct runs with account numbers of 1231xx where xx must be a string of two alphanumeric characters are considered for inclusion in the workload By using the amp amp operator only runs that meet both conditions are included in the workload Run_ID TQ amp amp Account 1231 Run_ID TQSYS Example 3 The following example shows a workload definition using account numbers to identify system jobs This workload contains information for all ru
97. same transaction code Total number of stuck transactions in the whole system Total number of RTPS deact copies with same transaction code Total number of RTPS deactivated transactions in the whole system Number of transactions active with same transaction code Total number of transactions active in the whole system Total number of transactions scheduled in the whole system Total number of transactions completed in the whole system Total number of user to user passoff messages Total number of output messages sent A 41 OS 2200 Parameters FCSS Read Func Reqs Total number of FCSS read function requests FCSS Write Func Reqs Total number of FCSS write function requests FCSS Lock Func Reqs Total number of FCSS lock function requests FCSS Other Func Reqs Total number of FCSS other function requests FCSS Read Words Total number of words read by FCSS read requests FCSS Write Words Total number of words written by FCSS write requests FCSS Locks Acquired Total number of locks acquired Exec Reads Total number of Exec reads Exec Writes Total number of Exec writes TIP Reads Total number of TIP reads TIP Writes Total number of TIP writes Transaction Count Total number of transactions T O Requests Total number of I O requests for this transaction Txns sec System Average number of transactions per second for the whole system A 42 TQ 01011 18 OS 2200 Parameters A 4 MIPS Parameters This subsection contains a l
98. sec IP SUPs sec Response sec run Throughput runs hour Population open runs Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec run Throughput runs hour Population open runs TQ 01011 18 System Log Overall Batch WLS1 WLS K words of memory per batch run for XPA systems Batch tape I Os per second Batch disk I Os per second Batch I O SUPs per second Batch ER CC SUPs per second Batch IP SUPs per second Seconds per batch run Batch runs per hour Average batch runs open System Log Workload Batch WLS1 WLS2 all K words of memory per batch run for XPA systems Batch tape I Os per second Batch disk I Os per second Batch I O SUPs per second Batch ER CC SUPs per second Batch IP SUPs per second Seconds per batch Batch runs per hour Average batch runs open A 57 OS 2200 Parameters A Y A 9 1 A 58 System Activity Monitor Statistics This subsection contains a list of the data items collected by the TeamQuest Baseline system activity monitor probe tqsam The information collected by the probe is stored in two tables e Runs Table e Transaction Table Note These data items are not collected by Team Quest Online Runs Table The following information is stored in the performance database tables Table Field Hierarchy Class 220
99. shown Brackets are used to indicate optional entries In the following command d databasename that which appears between the brackets d databasename is optional You are not required to enter the d parameter Italics are used to indicate variable values that you must supply depending on the circumstances In the following command a aggsetl you must supply a valid aggregation set name in place of aggset1 For menu selections within the graphical user interfaces a command sequence is shown as Administration gt Workloads This sequence indicates that to perform the necessary function you select Workloads from the Administration menu TQ 01011 18 Contents About This Guide Part 1 Getting Up and Running Section 1 Introduction N What Is TeamQuest Performance Software _ What Can You Do with TeamQuest Performance SOMWALC2 Arcache A E dteistavssareveunta aan fess What Is TeamQuest Baseline uu ceeeeeeeeseeeeeeeeeeeees What Can You Do with TeamQuest Baseline and TeamQuest Online occ cecceeeseseceeeeeeceeeeeeeeeeeeeeneenes Processing Overview ceeeececeeeeeeeeeeeeeeeeeeseteeeeeaeeeeaes Data Goll ction riras a a Data Display aa a ear aa ERARE AAAA ARN E How Do TeamQuest Baseline and TeamQuest Online WOLK A E EE AE E A A EE The Host Software Component neess ProDOS Oae a aa A aai Performance Database ccceeeeeeeeeeeeeees Host to TeamQuest View Interface _
100. statistics listed in the Derived Statistic Maintenance dialog box See 8 4 1 for an example of a derived statistics definition file TQ 01011 18 8 15 Customizing Derived Statistics 8 3 8 Displaying or Modifying a Derived Statistic Definition To display or modify a derived statistic definition select the statistic from the Derived Statistics list box in the Derived Statistic Maintenance dialog box see 8 3 1 and click Open You can also double click the statistic in the Derived Statistics list box The Derived Statistic Definition dialog box is displayed The definition information for the selected derived statistic is displayed in the dialog box 4 Derived Statistic Definition Session cl01 Database PRODUCTION Statistic Info jio SUPs sec Save As per Alias List iosups SYSTEM HYSTAT LogWorkload T otal 0 SUPs null WLS WL null elapsed SYSTEM System null null Elapsed Time null null null rull Function 10 SUPs sec iosups elapsed Using this dialog box you can modify the Statistic Info the Alias List and the Formula edit box For a description of these fields see 8 3 3 When you click OK your changes are accepted and you return to the Derived Statistics Maintenance dialog box see 8 3 1 Your changes are not actually made in the database until you click OK in the Derived Statistics Maintenance dialog box 8 16 TQ 01011 18 Customizing Derived Statistics 8 3 9
101. summary reports provide statistics which give a general overview of system performance These reports are found in the summary subdirectory of the sip subdirectory File Name Description File Name Description File Name Description File Name Description File Name Description TQ 01011 18 PR UTIL RPT Processor Utilization Reports processor utilization for user categories DEM RESP RPT Demand Response Reports demand response time LC CONCR RPT LC Concurrency Reports logical channel concurrency for each channel selected DSK ACT RPT Disk Activity Reports disk information for each selected disk IO SUM RPT I O Summary Summarizes I O rates for the system Predefined Report Descriptions File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description LC IOS RPT Logical Channel I Os Reports I O counts for the selected logical channels OPEN RUN RPT Runs Open Reports demand and batch population information TIP SUM RPT TIP Summary Shows total TIP transaction schedule and completion information PC ACT RPT Physical Chan Activity Reports physical channel information for each logical channel selected LC ACT RPT
102. table revision and the name of the file that contains the tables It is possible for a file to contain more than one table Administering Collection Periods Collection periods are used to control when data is collected for a given table With the tqzip program you can e List the currently defined collection periods e Add a collection period e Specify that individual days of the week are excluded from the collection e Specify how often the data values are obtained if the table contains intervalized data that is data where the values are obtained at regular intervals e Turn off data collection completely for a table while retaining the collection periods already defined To begin the administration of the collection periods for tables select Collection Periods from the Tables Menu see 11 3 by entering 2 at the prompt The tqzip program lists the tables defined in the database and asks you for the name of the table whose collections you wish to administer TQ 01011 18 Customizing Tables gt Enter your selection lt 1 gt 2200 Runs Alarm Alarm Log 2200 Transactions ter Table Number lt enter by name gt gt Class gt 2200 gt Name gt Runs The list of tables depends on the particular types of data being stored in tables Enter the number of the table whose collections you wish to administer If you do not enter anything tqzip asks you for the complete table name including the table class and table nam
103. tablefile gt tqzip u c tablefile gt TQZIP 7R4G 011312 0704 45 Fri Jan 13 08 11 27 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt Using database TQDATA PRODUCTION gt gt unloading table specifications to tablefile gt gt Table information unloaded gt gt End TQZIP TQ 01011 18 11 21 Customizing Tables Table Specification File Format The table specifications appear in the following format in the output file created by tqzip If you choose to edit this file or update the definitions you should maintain this file format See 11 5 and 11 6 for details on adding or modifying collection and retention specifications BEGIN TABLESPECS unloaded on Fri Jan 13 14 33 57 2012 A 2200 Runs RetentionStart 0000 PrimaryRetention 1DAY RetentionSpec 001000 5DAY w RetentionSpec 010000 1MONTH Collection 0N CollectionSpec 0000 2400 0030 0000000 FieldSpec 0rig_Runid ON FieldSpec Account ON FieldSpec Program ON FieldSpec Pgm_Start ON FieldSpec KSize ON FieldSpec Run_Elap ON FieldSpec Int_Elap ON FieldSpec Tot_Sups ON FieldSpec Int_Sups ON FieldSpec Int_ Sups 0ON FieldSpec Tot_IP 0 FieldSpec Int_IP 0 FieldSpec Tot_I0 0 FieldSpec Int_I0 0 FieldSpec Tot_ER CC ON FieldSpec Int_ER CC ON FieldSpec Tot_I0req ON FieldSpec Int_I0req ON FieldSpec Tot_IOrate ON FieldSpec Int_IOrate ON FieldSpec Sample_Time ON FieldSpec Type ON Fiel
104. text gt 8 User text gt Modify the thresholds y lt n gt D gt gt Threshold entry created TQ 01011 18 7 11 Customizing Alarms and Thresholds The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A To reuse a threshold definition for a resource enter y at the prompt The tqzip program returns you to the list of resources see E in 7 4 B Enter the new resource for which you want to define a threshold You can enter either the number of the resource or the resource name c The tqzip program displays the threshold definition for the last resource defined D The tqzip program allows you to use the threshold definition as is or to modify it for the current resource If you enter n at the prompt the new threshold definition is saved to the performance database immediately If you answer y at the prompt the tqzip program displays the series of threshold definition prompts with the defaults set to those displayed in the definition see G through P in 7 4 7 6 Updating Threshold Definitions The tqzip program allows you to update threshold definitions that are currently defined in the performance database To update a threshold definition select Define Thresholds from the Alarms and Threshold Menu see 7 3 by entering 1 at the prompt The tqzip program displays the following prompts to update thresholds for a pa
105. that exist in your database The derived statistic names are made up of the Category Group Category Subcategory and Statistic keys To report on derived statistics you enter these four keys as part of a parameter_selection definition just as you do when reporting stored parameters In addition you must enter a system name and you may need to include the resource workload set workload and application group names on the parameter_selection definition for the derived statistic When you define a derived statistic formula aliases are used to refer to other parameters in the database Substitution indicators are used in aliases to indicate that certain key values are to be specified at report time For example you may use a substitution indicator for the resource name key of a given alias To report on the derived statistic you would specify the resource as part of the parameter_selection definition entered at the tqrpt prompt For more information on derived statistic referencing aliases and substitution indicators see 2 6 TQ 01011 18 D 5 Host Based Reporting for Aggregation Set Data Example To illustrate the use of aliases perhaps you have defined a derived statistic in your database as follows Derived statistic Derived Derived Category Derived Subcategory processory busy name Aliases idle SYSTEM Processor Processor Utilization Idle RES Formula 100 idle To report on the derived statistic processor busy you
106. the h option usage information containing the command line syntax is printed and the program then terminates Parameters passed to a program from the Exec call line are in lowercase and must have valid file name syntax Some parameters on the tqtblprb call line are case sensitive or have characters that are invalid for a file name These parameters must appear on the second and subsequent lines of the processor call Command line option values should be enclosed in quotation marks if the value contains spaces Table Creation The syntax for executing tqtblprb to create a table is filename tqtblprb i f file d databasename end where i specifies a table creation execution f specifies the name of a file that contains table specifications for creating the table If the f option is not specified the table specifications are read from the runstream d specifies the database into which data is stored The databasename parameter is the name of an existing database on your system If the databasename is not specified the database in the tq ini file is used The following shows the table creation command for the example The tqtblprb program will read the usertabs file for creating the table filename tqtblprb i f usertabs end Note The tqtblprb program sets T3 of the run s condition word to 1 to indicate the table already exists It sets T3 to 2 for all other error terminations This could be used in runstreams to
107. the z option you must specify the t option The z option can be used to create an initial report definition suitable for editing TQ 01011 18 Host Based Reporting for Table Data n prints the specified report for the table specified by the t option The defRepname parameter is the name of a report that has been defined for the table If you use the n option you must specify the t option To find a list of the available report names for a table use the l option f prints the report from the specified report definition file The reportdefnfile parameter specifies the name of a file that contains a report definition 0 prints the report to the specified output file If this option is not specified the output file specified in the report definition is used X exports the data using the specified delimiter The delimiter parameter can be any printable character string In addition the reserved words TAB COMMA and SPACE can be used to indicate a tab comma or space delimited report respectively W specifies the number of characters to use for the report width Any text outside the specified width is not printed Note You must request some type of reporting for each tqrep execution The t l or f option must be specified to indicate the type of reporting to be performed Example The following command was used to generate the report in Example E 1 The f option specifies the report definition file alarm re
108. the 30 minute interval records that were aggregated The 1 hour records are retained for 6 months At the end of the 6 months the records are removed from the table For this example data records are retained for a total of 7 months 1 month of 30 minute interval data and 6 months of 1 hour data 1 week 1 minute interval data and 1 day data at the interval collected 11 6 2 Modifying the Primary Retention Specification 11 12 The primary retention specification specifies how long the records created by the data collector will be retained The interval of the records remains the same as when the data collector inserted them Only the retention period can be modified for the primary retention specification To modify the primary retention period specification select Modify Primary Retention Specification from the Retention Specification Menu see 11 6 by entering 2 at the prompt The tqzip program displays the following prompt gt Enter your selection lt 1 gt gt 2 gt gt Retention period lt cancel gt gt 2 days At the prompt enter a number and a unit for the retention period Valid units are minute hour day week month and year In the example the retention period for which the data collector will create records is changed to 2 days TQ 01011 18 Customizing Tables 11 6 3 Adding a Retention Specification The tqzip program allows you to add a retention specification to a table When you add a rete
109. the characters in which the x axis title is printed The value must be a positive integer C 14 TQ 01011 18 Report Definition File Foreground Indicates the color in which the x axis title is printed Background Indicates the color of the area that appears directly behind the x axis title of the axis The background area is a boxed area and acts as a frame for the x axis title If a background color is specified the boxed area behind the x axis title covers any graph lines and other text that may appear behind the x axis title If the Background field for the x axis title is not specified TeamQuest View allows any graph lines or other text that may be behind the x axis title to show through Border Indicates the color of the border line that appears for the background area of the x axis title If the Border field is not specified TeamQuest View does not display a border line around the background area of the x axis title The default values for XAxis Style are defined in the Report Styles settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences XAxis Line Enable Thick MajorTick MinorTick MajorGrid MinorGrid TQ 01011 18 The specifications for the line attributes of the horizontal axis of the report Enable Indicates whether the horizontal axis line appears on the report The acceptable values are ON or OFF If Enable is set to ON the horizontal axis line appears on th
110. the data in near real time mode tqtpm The tqtpm probe processes TIP HVTIP statistics from the TIP Performance Monitor TPM audit trails created when TIP logging is configured This probe requires that your TPM audit trail be configured to mass storage or moved to mass storage prior to starting either probe The tqtpm probe automatically processes the TPM audit trail s from the previous day from TPM audit trail s specified on the call statement or based on a relative time selection also specified on the call line The tqtpm probe extracts statistical values plus date and time information from the TPM audit trail s determines the workload that used the resource and stores this information in the performance database according to the workloads defined for the system tqiot The tqiot probe processes statistics from the IOTRACE data collection file The tqiot probe processes the IOTRACE data files specified on its processor call statement The tqiot probe extracts statistical values on system resource utilization plus date and time information from the data file determines the workload that used the resource and stores this information in the performance database according to the workloads defined for the system tqmpr The tqmpr probe processes data from the MAPPER accounting log files The tqmpr probe processes the set of MAPPER accounting log files from the previous day or the MAPPER accounting log files specified on its processor ca
111. threshold definitions with tqzip use the following processor call format tqzip d database u t threshold file where d specifies the database that tqzip uses The database parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 u specifies that an unload operation is to be performed by tqzip The type of unload operation is thresholds as indicated by the t option t unloads the threshold definitions from the database to the file specified by the threshold file parameter TQ 01011 18 7 21 Customizing Alarms and Thresholds Example The following shows an example of unloading threshold definitions to the file tfile gt tqzip u t tfile gt TQZIP 7R4G 011312 11 04 45 Fri Jan 13 15 37 08 2012 gt Copyright c 1992 2012 TeamQuest Corporation gt All Rights Reserved TA database TQDATA PRODUCTION E ca ihce thresholds to tfile F threshold definitions were unloaded i TQZIP Threshold Definition File Format The threshold definitions appear in the following format in the output file created by tqzip If you update the threshold definition file you should ensure that the format of the file remains consistent with what was output by the u and t options The following example shows how the definitions appear in the file BEGIN THRESHOLDS unloaded on Fri Jan 13 10 15 4
112. time Average hardware service time Concurrency Unit average response time in milliseconds K words transferred per second Average number of requests queued to this unit Channel queuing time in milliseconds IO Activity Tape TAPO TAP1 Summary Tape percent busy Number of requests per second TQ 01011 18 OS 2200 Parameters Percent Reqs Read Percent Reqs Queued KWords Req KWords Req In KWords Req Out Transfer Time RET FLD Hardware Serv Time Concurrency Response Time KWords Sec Avg Queue Length Channel Queue Time Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec Percent Reqs Read Percent Reqs Queued KWords Req KWords Req In KWords Req Out Transfer Time RET FLD Hardware Serv Time Concurrency Category Group Category Statistic Name Destage Reqs Sec Destage Avoided Speculation Success Disk Traffic Hit Rate TQ 01011 18 Percentage of all requests that were reads Percentage of all requests that were queued Average number of K words per request Average number of K words per request that were input Average number of K words per request that were output Average transfer time per request Average service time Average hardware service time Concurrency Unit average response time in milliseconds K words transferred per second Average number of requests queued to this unit Channel queuing time in milliseconds IO Activity Other DCPO
113. transactions to run the probe in real time to specify the database name and to specify the maximum number of longest running transactions to monitor The tqgsam runstream can be edited and tailored to your site specific needs The tqsam processor call has the following format tqsam r t d database name n transactions END where r executes tqsam in real time mode The account for the tqgsam run must have a maximum real time level greater than zero By running in real time mode you can prevent the probe from missing data as a result of being swapped out during scheduled sample periods The default value is to not run in real time mode t specifies that statistics on transactions should be collected If you are not running TIP or do not want TIP statistics collected do not use the t option This will reduce processing time and database storage overhead The default value is not to collect TIP statistics TQ 01011 18 9 5 Customizing Data Collection Probes 9 4 2 d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini file is used Make sure that all programs called in a runstream use the same database name n specifies the number of active longest running transactions that should be monitored stored and reported in the table report This
114. use the following steps to define a new session 1 Select File gt Session gt New from the menu bar 2 Inthe New Session dialog box specify Socket as the connection type The Host Name or IP address of your OS 2200 system The Port number in use by TQTCP on your OS 2200 system When you click OK in the New Session dialog box the connection is made to your OS 2200 system and a TeamQuest View or a TeamQuest Performance Software Administration session is initialized TQ 01011 18 I 11 Appendix J 2200 SNMP Trap Alarms J 1 Overview The TeamQuest Performance Software provides the capability of sending a Simple Network Management Protocol SNMP trap as an alarm action Using the tqzip program you can define thresholds that will send an alarm event to any management console that accepts SNMP traps Most management consoles allow you to specify a variety of actions upon receipt of an SNMP trap such as making an entry in an event window popping up an alarm box window or even running a program such as TeamQuest View The TeamQuest Baseline uses the SNMP TSAM interface supplied by the Unisys product CMS 1100 to send SNMP traps when an alarm occurs on a managed node The tqalm program of TeamQuest Baseline continuously runs to monitor for alarm conditions on the system In CMS 1100 network management terms tqalm is an SNMP Management Agent Extension that sends SNMP traps to the CMS SNMP agent The CMS SNMP agent then forwards th
115. used see 3 1 C causes the tqagg program to run in continuous mode In continuous mode tqagg performs the data consolidation each day at midnight If you want to run tqagg in continuous mode you should update the tqagg runstream appropriately If the c option or t option is not specified tqagg performs data consolidation for each fed aggregation set and then terminates TQ 01011 18 6 27 Customizing Aggregation Sets 6 28 S f Notes causes the tqagg program to run in continuous mode This option differs from the c option in that you can specify the time you want the data consolidation to take place each day The time is specified as hhmm If you want tqagg to run in continuous mode you should update the tqagg runstream appropriately If the c option or t option is not specified tqagg performs data consolidation for each fed aggregation set and then terminates allows you to reprocess a time period that has been consolidated previously With the r option you can specify the number of days back to begin the consolidation process For example if you specify 2 days the tqagg program starts aggregating the data beginning two days prior to the last point it had consolidated Previously consolidated data points remain unchanged but any data points stored to the source aggregation set since the last tqagg execution are now consolidated Without the r option tqagg begins the data consolidation for an aggregation
116. used with the p option must be in the same format as the file produced by the o option see 10 8 10 6 TQ 01011 18 Customizing Performance Databases Example To delete database parameters generate a parameter list file using the o option of tqzip see 10 9 Then edit the file The following example shows the contents of a parameter file pfile BEGIN PARAMETER LIST created from lt PRODUCTION gt on Fri Jan 13 14 05 45 2012 System CatGroup Cat SubCat Stat Res App WL Set WL CLO1 10 Activity Disk Summary DSKF END PARAMETER LIST This shows that you want to delete all statistics for DSKF7 You then call tqzip with the p option and specify pfile on the processor call gt tqzip p pfile gt TQZIP 7R4G 011312 1120 33 Fri Jan 13 14 50 17 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt gt Using database TQDATA PRODUCTION gt gt generating list of pnums to delete gt 14 parameters selected for deletion gt Proceed with deletion y lt n gt y gt gt deleting parameters gt gt 14 parameters were deleted gt gt End TQZIP 10 10 Database Locking The aggregation set database files that is those files named TQDATA xxxxaaaaaaaa are locked by each probe during an update When one probe has an aggregation set file locked all other probes will wait until it becomes available before storing their data If a probe terminates abnormally its contingency
117. using SYS LIB RUN TQCTL The tqagg runstream is started with a start time of midnight The tqagg run remains in backlog until midnight After it finishes processing tqagg puts itself into backlog again until the next midnight The c and t options provide an alternative method of controlling when tqagg is run TQ 01011 18 Section 7 Customizing Alarms and Thresholds 7 1 7 2 Note The alarms and thresholds capability is available with TeamQuest Baseline only Introduction The tqzip program is used to administer the alarms and thresholds capability of the performance database Default thresholds are supplied and can be loaded into your performance database see 3 3 2 If you are content with the default threshold values no additional administration is required However if you want to customize your thresholds you can perform the following tasks e Define thresholds see 7 4 7 5 and 7 6 e Define thresholds for derived statistics see 7 8 and 7 9 e View threshold information see 7 7 and 7 10 e Delete thresholds see 7 11 e Notify the probes of changes in thresholds see 7 12 e Unload thresholds see 7 13 e Load thresholds see 7 14 e Turn on and off threshold checking see 7 15 e Create a file containing alarm messages see 7 16 Concepts The TeamQuest Baseline probes collect data from various data sources at the sample interval defined for the user specified aggregation set The probes can also check
118. values should only be specified when new default file names are required The parameter should be removed for normal execution of the probe It should not be used with the f parameter see 9 9 3 TQ 01011 18 9 9 5 9 10 Customizing Data Collection Probes Workload Characteristics for the MAPPER Log Probe The MAPPER log probe has a unique set of characteristics that are used for its workload definitions You use these characteristics when you create the workload definitions using TeamQuest Performance Software Administration see 5 3 The following characteristics define the MAPPER log workloads Characteristic Valid Type and Valid Name Characters Max Length Operations User A Z 0 9 string 12 l Dept A Z 0 9 string 11 Se l Function A Z 0 9 string 12 Run A Z 0 9 string 12 ss Type 0 9 integer 6 lt 55 lt gt Station 0 9 integer 12 lt gt lt gt Example The following example shows a workload specification using MAPPER department names to identify Overhead interactions This workload will contain information for department names PAYROLL and PAYROLLx MIS and SYSTEM By using the Dept PAYROLL Dept PAYROLL construct data for interactions with department identifiers of PAYROLL and PAYROLLx where x can be any single alphanumeric character will be included in this workload Dept PAYROLL Dept PAYROLL Dept MIS Dept SYST
119. verification C If you enter y at the prompt the tqasa program will continue and prompt you for the name of the aggregation set to convert Enter n if you have not backed up your database or if there are processes currently running that are using the selected database D Enter the name of the aggregation set that you want to convert from expanded to compact format E The tqasa program performs the conversion of the aggregation set and displays a message when the conversion is complete Performing Data Aggregation When you create an aggregation set that is fed by another aggregation set you must execute the tqagg program so that data is retrieved from the source aggregation set consolidated and stored into the fed aggregation set The tqagg program performs data aggregations for all fed aggregation sets defined in a performance database The tqagg program is also responsible for consolidating and removing table records The runstream SYS LIB RUN tqagg is started to run tqagg By default this runstream is set up to run at midnight The tqagg program has the following processor call format tqagg d databasename c t hhmm r days s aggsetname f aggsetname where d specifies the database for which you want to perform data consolidation The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is
120. was recorded in the database for this sample Interval Double The desired sample interval in this collection period Client String The name of the client which is the tp_apname specified in the AP NAME element of the client program Gen_Runid String The generated OS 2200 run id Service Table Information The following information is stored in the performance database tables Table Field Hierarchy Class 2200 Name ODTP Service File name odtpfile Collection interval Based on the table collection interval Default retention 1 day Data Item Type Description Time Double The time this information was recorded in the database for this sample Interval Double The desired sample interval in this collection period Service String The name of this service Type String The service type Local Remote HAA or TWRAP Server String The configured name of the server or TRANCODE for an HAA service or TWRAP service TQ 01011 18 OS 2200 Parameters Fails Double The number of service requests that failed in this sample interval Aborts Double The number of service requests processed that aborted in this sample interval Sleeps Double The number of times waiting for messages from a a call or a conversation in this sample interval Sleep_Time Double The time spent waiting for messages from a call or a conversation in this sample interval Receives Double The number of conversational receives issued in this sample interval Sends D
121. where x can be any single alphanumeric character is included in this workload Run_ID MSCP Run_ID MSCP Run_ID RUNNE Run_ID CMR Run_ID CMR Run_ID PCCTI Run_ID PCCTI Run_ID DDP Run_ID DDP Example 2 The following example shows a workload definition using file name identifiers to group all I O activity to a set of specific files File ONL XYZ DIAG RLIB TQ 01011 18 Customizing Data Collection Probes 9 9 MAPPER Log Data Collection Note The MAPPER log probe is released with TeamQuest Baseline only It is not available with TeamQuest Online The MAPPER log probe uses data from MAPPER accounting log files The MAPPER software automatically creates this accounting log file At various stages in the life of a MAPPER run or manual function entries are made into this log The entry includes information about the resources used by the function The site can cause the MAPPER software to create additional log entries for diagnostic purposes This is known as turning on run function logging With run function logging off an entry is made for each manual function for each run and additional entries for displays and for certain other functions With run function logging on an entry is made for each manual function for each run and for each function within a run for which there is a corresponding manual function Run function logging can be turned on globally for all
122. 0 Name Runs File name samfile Collection interval Based on the table collection interval Default retention 1 day at primary retention specification Data Item Type Description Run_ID String The unique generated run id of the run Orig _Runid String The original run id of the run Tot_Sups Double Cumulative SUPs for this run since inception Int_Sups Double SUPs for this run in this interval Int_ Sups Double Percentage of total SUPs used by this run in this interval Program String The name of the program the run is executing KSize Integer The current size in K words of the run Account String The account id of the run Type String Type of run It can be one of the following demand demand demand rt demand real time batch batch batch rt batch real time tip tip Time Double The time this information was recorded in the database for this sample Sample_Time Double The actual time that this sample was taken for the run TQ 01011 18 OS 2200 Parameters A 9 2 Interval System Run_Start Run_Elap Int_Elap Pgm_Start Tot_IP Int_IP Tot_IO Int_IO Tot_ER CC Int_ER CC Tot_IOreq Int_IOreq Tot_IOrate Int_IOrate Count Double String Double Double Double Double Double Double Double Double Double Double Double Double Double Double Integer Transaction Table The desired sample interval in this collection period The system from which this sample was taken The time the
123. 1 fields of data that describe the alarm and the default is to send all of these fields in the trap However you can also choose to send only selected fields depending on your site requirements TQ 01011 18 2200 SNMP Trap Alarms SNMP Trap Syntax You specify an SNMP trap alarm action by placing a special syntax on any one or more of the User Text fields in a threshold definition see 7 4 The syntax for this command is as follows SNMPTrap Fields field list The field list is an optional list of trap variables to send to the trap destination It consists of any of the variable names described in J 4 2 The variable names in the list must be separated by a comma If the Fields parameter is not specified on the command the default is to send all variables The following rules apply when entering the SNMP trap syntax on the User Text fields in a threshold definition e The SNMPTrap keyword must begin in the first character position of the field e The SNMPTrap and Fields keywords are case sensitive and must be separated by one or more spaces e The command can be continued to the next User Text field by placing a backslash as the last entry on the image Each User Text field can be up to 60 characters long e Only one SNMPTrap action can be specified for each threshold The User Text fields that contain this command will not be part of an alarm message sent to the console or become part of the trap s fields Exa
124. 10 seconds Zone 2 and remains at this level for 60 seconds an alarm message with a severity of CRITICAL is sent to the console The current zone for demand response time is changed to be Zone 2 The next alarm would be generated when the value for demand response time returned to Zone 1 for at least 60 seconds This alarm would have a severity of NORMAL indicating a return to the normal zone that is zone 1 TQ 01011 18 7 3 Customizing Alarms and Thresholds Opening the Alarms and Thresholds Menu The tqzip program displays a series of menus for you to select the tasks you want to perform To perform maintenance on alarm and threshold information you execute the tqzip program and select Alarms and Thresholds from the Administration Menu gt tqzip gt TOQZIP 7R4G 011312 11 04 45 Fri Jan 13 09 19 55 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt gt Using database TQDATA PRODUCTION gt gt gt Administration Menu gt gt 1 Alarms and Thresholds gt 2 Tables gt gt 99 Exit TQZIP gt gt gt Enter your selection gt 1 The tqzip program then displays the Alarms and Thresholds Menu Alarms and Thresholds Menu 1 Define Thresholds 2 View Thresholds 3 Delete Thresholds 4 Notify Probes of Threshold Changes 5 Turn Threshold Checking On Off 6 Generate Alarm List 99 Exit Alarms and Thresholds Menu Enter your selection MEN NTO
125. 1011 18 Double Double Double Double Integer Double Double Double Double Double The number of service calls issued to other services that did not use two phase commit in this sample interval The number of service requests processed that were not within a two phase commit in this sample interval The number of service calls issued to other services that used two phase commit in this sample interval The number of service requests processed that were within a two phase commit in this sample interval The number of combined occurrences This is collected when STATS LEVEL gt 2 Number of two phase commit transactions initiated by a service in this sample interval Average time for each two phase commit service transactions in this sample interval Number of transactions committed that were initiated by a service in this sample interval Number of transactions rolled back that were initiated by a service in this sample interval Number of transactions that timed out that were initiated by a service in this sample interval A 63 OS 2200 Parameters A 10 3 A 10 4 A 64 Client Table Information The following information is stored in the performance database tables Table Field Hierarchy Class 2200 Name ODTP Client File name odtpfile Collection interval Based on the table collection interval Default retention 1 day Data Item Type Description Time Double The time this information
126. 1011 18 Index host based reporting D 1 overview of processing 1 5 probe 1 8 tqdata 1 11 tq ini element 3 1 TeamQuest MIB definition for SNMP traps J 9 TeamQuest Online data collection 1 6 data display 1 6 host software description 1 8 overview of processing 1 5 probe 1 8 tqdata 1 11 tq ini element 3 1 TeamQuest Performance Software Administration additional sessions 4 6 aggregation set tasks 6 8 derived statistic tasks 8 4 establishing a session 4 1 introduction 4 1 socket connection 4 1 terminating 4 8 warning screens 4 7 workload tasks 5 8 TeamQuest View alarm messages displayed by 7 30 data display 1 6 description 1 11 predefined report descriptions B 1 retrieving data 1 6 terminating administration session 4 8 SIP data collection 9 3 System Activity Monitor data collection 9 6 tqbsp 9 3 tqsam 9 6 threshold definitions changing defaults 3 5 defaults 3 4 using defaults 3 4 thresholds checking 7 21 creating 7 6 creating for derived statistics 7 14 deleting 7 19 exporting definitions to a file 7 21 loading definitions 7 24 maintaining 7 1 reusing threshold definitions 7 11 TQ 01011 18 turning on and off threshold checking 7 25 unloading definitions 7 21 updating definitions 7 12 updating for derived statistics 7 16 viewing 7 13 TPM log data collection description 9 17 file input totqtpm 9 21 parameters A 41 reports B 17 starting 9 17 tqagg program command line format 6 27 desc
127. 11 18 OS 2200 Parameters RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec Percent Reqs Read Percent Reqs Queued KWords Req KWords Req In TQ 01011 18 Logical channel read bypass service time Logical channel read bypass hardware service time Logical channel read bypass percent of all requests IO Activity Logical Channel LCO LC1 Write Bypass Logical channel write bypass requests per second Logical channel write bypass K words per request Logical channel write bypass transfer time Logical channel write bypass service time Logical channel write bypass hardware service time Logical channel write bypass percent of all requests IO Activity Logical Channel LCO LC1 Store Through Logical channel store through requests per second Logical channel store through K words per request Logical channel store through transfer time Logical channel store through service time Logical channel store through hardware service time Logical channel store through percent of all requests IO Activity Control Unit CUO CU1 Summary Control unit
128. 11312 11 17 30 Fri Jan 13 08 17 27 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt removing all existing aggregation sets gt loading aggregation set definitions from aggsetfile gt File created TQDATA PROD1 minute gt File created TQDATA PROD10 minute gt 2 definitions were loaded gt gt Aggregation Set List gt gt Max gt File gt Name Gran Start Stop Keep Size Source Name gt SR ines bee feet Co Beene erties ete te ee Sea te ete Serie Cee ee gt l minute 00 01 00 00 00 00 23 59 59 08 00 00 fo 4 22 lt gt 10 minute 00 10 00 00 00 00 23 59 59 8 dayS gt gt gt An following Name indicates an expanded aggregation set gt gt End TQASA TQ 01011 18 6 25 Customizing Aggregation Sets 6 4 6 6 26 Converting Expanded Aggregation Sets The tqasa program allows you to convert an existing expanded aggregation set into compact format It is reeommended that you back up your performance database before performing this conversion Also all processes that are currently using the selected database must be terminated before running tqasa to convert an aggregation set To convert an existing aggregation set from expanded to compact format the tqasa processor call has the following format tqasa cL d databasename where C calls the tqasa program to begin aggregation set compaction d specifies the database th
129. 19 workloads definition file format 5 19 language 5 2 rules 5 2 description of concepts 2 4 exporting definitions toa file 5 19 IOTRACE probe 9 28 loading workload definitions 5 22 MAPPER log probe 9 33 specifying workload definitions 5 1 system log probe 9 37 TeamQuest Performance Software Administration 5 8 TPM log probe 9 23 tqzip administration 5 18 unloading workload definitions 5 19 workload definition language 5 1 workloads characteristics for IOTRACE probe 9 28 for MAPPER log probe 9 33 for System Activity Monitor probe 9 7 for system log probe 9 37 for TPM log probe 9 23 workstation software description 1 11 X XPC statistics A 36 Index 10 TQ 01011 18
130. 30 minute intervals 1 hour intervals or perhaps 8 hour intervals Each data value in an aggregation set represents the value for a parameter over the specified time interval e A type of date consolidation Specify the type of date consolidation represented by each point maintained in the database for the aggregation set TeamQuest Baseline supports date consolidation types of day week and month For example you can have an aggregation set that stores data records with 30 minute granularity throughout each day A single data point exists for each 30 minute interval for each day the data was collected Another aggregation set can be defined that still has a granularity of 30 minutes but where each data point represents a 1 week or 1 month average for each 30 minute interval For the week and month date consolidation types you can choose days of the week to exclude from the aggregation set For the week date consolidation type you can also choose the day of the week that is considered to be the first day of the week e The retention period for the data Specify the amount of time to maintain the data in the database for the aggregation set For example you could retain your data for one week two months or one year The retention period determines the amount of data in your database The longer you retain your data the larger your database will become e The start and end time of data sampling Specify the time range for sampling data
131. 38338 340 1397 323 80 12044 OVERFLOW 3455 6140518 38 1738 157 97 2529 TOTAL 4455 7878856 106 1661 250 93 14573 TQ 01011 18 9 43 Customizing Data Collection Probes An SSG skeleton can be written that inputs the DMU output as SGSs The skeleton can then loop through the OVERFLOW SGSs and output the information in a format that tqusrprb can read This skeleton creates lines of data that are in the format described by the TQARG file For this example the TQARG file is as follows RESOURCE pages Total words NONE NONE NONE Unused Total Records NONE AVG AVG NONE NONE NONE AVG AVG The SSG skeleton creates a PRBDATA file This file is read by tqusrprb and the statistics are inserted into the performance database The SSG skeleton is as follows output the OVERFLOW information BRKPT prbdata SET X INCREMENT i to OVERFLOW i F COVERFLOW i lt 7 CSTATISTICS xJ 3 1 0000000 x ELSE CSTATISTICS x 3 1 OVERFLOW 7 1 1 6 99 99 ENDIF SSE X AXI LOOP The resultant PRBDATA is as follows PRIMUS D01 3455 6140518 38 1738 157 97 2529 A runstream is created that assigns the TQARG and PRBDATA files Then the SSG skeleton is called with the breakpointed output of DMU as input SGSs The tqusrprb program is then executed to insert the data into the performance database gt RUN TQUSR accnt TQDATA gt ASG A TQARG gt ASG T PRBDATA gt SSG A DMUskel DMUbrkpt gt EOF gt
132. 4 10 5 10 6 10 7 10 8 10 9 10 10 10 11 ItKOCUGTION deeccheateevbecsssheeaceeeleaeezs TAT Performance Databases ou ccceceeeeeeeeeenteceeeeeeeeeeteees Database Files aa ae fee tivities Creating Additional Databases eceesecceeeesseeeees Deleting Databases o eeeecccceesecceceeeeeecceeeeeseeeeeeeeetenes Listing Existing Databases oo ceeeteeeeeeenteeeeeees Estimating Disk Space Requirements _ ccceeceees Listing Database Parameters ecccceeesseeeeeeeetteeeeeees Removing Database Parameters from a Database _ Database LOCKING seresa ran EANA Backup and Recovery ssessssssssssrresssrnnesrnnnessrnnnnensrenne Section 11 Customizing Tables 11 1 11 2 11 2 1 11 2 2 11 2 3 11 2 4 11 3 11 4 11 5 11 5 1 11 5 2 11 5 3 11 5 4 19 5 11 6 11 6 1 11 6 2 11 6 3 11 6 4 11 6 5 11 6 6 11 7 11 7 1 11 7 2 11 7 3 11 8 11 9 11 10 11 11 11 12 11 12 1 11 12 2 xvi IMMFOCUCHION Eaa EEE keeles oh A GONCEPIS Pasan ar aaa earan aa aaraa EEEa Saaana Collections Lora a E D RELENtlOnS ives r r oii ann aaaea a a asai Registration A senri sarene ara A TA Manual Record Deletion cccceeccceeeeseeeeeeeeeaes Opening the Table Administration Menu eeee Displaying a List of Tables oe eee ceeeeeeeeeneeeeeeeeeaee Administering Collection Periods sccceecsseeeeeeeeees Listing Collection Periods ccs
133. 6 months gt Continue Notice that the 15 minute retention falls between the 1 minute and 30 minute retentions The data is retained at 1 minute intervals for 1 week then at 15 minute intervals for another week then at 30 minute intervals for 1 month and at 1 hour intervals for 6 months TQ 01011 18 Ta Customizing Tables 11 6 4 Deleting a Retention Specification You can delete any retention specifications for a table except the primary retention The primary retention specification must always exist Deleting a retention specification does not result in the immediate deletion of the data in that retention period The next time the aggregator processes the data in the table the data records will be aggregated so that they match the new retention specifications To delete a retention specification for a table select Delete a Retention Specification by entering 4 at the prompt The tqzip program displays the following prompts gt Enter your selection lt 1 gt gt 4 gt gt Retention to Delete lt cancel gt A gt gt Retention Specifications 5 B gt gt Number Interval Keep gt Bp cia ahaa tema OARA dN ae cad tab aa es a aaa rae gt Primary 9 gt 2 days gt 1 00 01 00 1 week gt 2 00 15 00 1 week gt 3 00 30 00 1 month gt 4 01 00 00 6 months ARetention to Delete lt cancel gt C Al The following lettered descriptions correspond to the capital letters
134. 75 gt Value must be in new zone for lt 0000 gt seconds before raising alarm L gt 60 gt Disable threshold checking for lt 0000 gt seconds after raising alarm M gt 60 gt gt To clear a text string enter a period at the prompt gt gt Alarm identifier lt gt N gt Percent Busy gt User text line 1 lt gt 0 gt When more than 50 busy call John at ext 2332 gt User text line 2 lt gt gt gt User text line 3 lt gt gt gt TQ 01011 18 7 7 Customizing Alarms and Thresholds gt CLO1 10 Activity Disk Summary Percent Busy DSKFl P gt 1 Threshold checking ON gt 2 Threshold 1 00050 00 Threshold 2 00075 00 gt 3 Number of zones EO Normal Zone ee gt 4 Secs before alarm 0060 Secs delay 0060 gt 5 Alarm identifier Percent Busy gt 6 User text When more than 50 busy call John at ext 2332 gt 7 User text gt 8 User text gt Save this entry lt y gt n gt y gt gt Do you want to define another threshold y lt n gt Q gt n gt gt Threshold checking is ON R gt Notify active probes of these threshold changes y lt n gt gt y gt gt Probe notification complete The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A Enter c at the prompt to create a new threshold definition The tqzip program then displays a series of pro
135. 8 Fed_By END AGGSET N The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The first line of the aggregation set definition file must be the header line BEGIN AGGSETS It is followed by a series of aggregation set definitions This line begins the first aggregation set definition Each aggregation set definition is made up of 10 lines The lines must appear in order and the label must appear exactly as shown followed by an equals sign The lines are as follows Name Granularity Start_Time Stop_Time The aggregation set name A value for this field is required The granularity of the aggregation set specified in seconds For example if the aggregation set contains hourly data the granularity is 3600 A value for this field is required The time to start data collection for the aggregation set It is specified as hhmm ss If no value is specified for this field 000000 is used by default The time to stop data collection for this aggregation set It is specified as hhmm ss If no value is specified for this field 235959 is used by default TQ 01011 18 Customizing Aggregation Sets Group_Type The group type of the aggregation set This value can be DAY WEEK or MONTH If no value is specified for this field DAY is used by default Week_Start The day that is considered the first day of the week fo
136. 9 2012 A DERIVED Derived Derived Category Derived Subcategory UserCpuUtil B 1 Threshold checking ON C 2 Threshold 1 Si 00 Threshold 2 00000 00 3 Number of zones Normal zone 1 4 Secs before alarm i Secs delay 0000 5 Alarm identifier UserCpuUtil 6 User text Call John at extension 2332 7 User text 8 User text 9 System Name Count 1 D CLO1 10 Resource Name Count 2 IP000 IP001 11 Workload Name Count 0 12 Application Group Name Count CLO1 10 Activity Disk Summary Percent Busy DSKF1 E Threshold checking ON F 2 Threshold 1 00050 00 Threshold 2 00075 00 3 Number of zones IS ormal zone e ol 4 Secs before alarm 0060 Secs delay 0060 5 Alarm identifier Percent Busy 6 User text When more than 50 busy call John at extension 2332 7 User text 8 User text END THRESHOLDS G 7 22 TQ 01011 18 Customizing Alarms and Thresholds The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A TQ 01011 18 The unload output file starts with a BEGIN THRESHOLDS line This line must not be changed or deleted This is the first line of a threshold definition for a derived statistic Each derived statistic threshold definition consists of a derived statistic name line followed by eight lines of descriptive information and four lines for qualifier values The derived statistic name line m
137. AL 223 ca Oa Se tq2200AlarmDateTime 1 36 LAL 223 05 641 6 tq2200AlarmText1 W136i de ALT 22307 208 tq2200AlarmText2 Te Sc6e ds cl a223274 61 8 tq2200AlarmText3 Te 366 TAS 3223 7 269 tq2200AlarmDatabase TBO AL 223 7504 1 10 tq2200AlarmParameter A oO vd eA L223 273601 e 1 4 Update the tecad_snmp cds file in use by the SNMP Event Adapter by appending the entries contained in the element SYS LIB TQ BASELINE TECAD CDS The tecad_snmp cds file describes the mapping between SNMP traps and the events that are forwarded to the TEC Event Server 5 Update the tecad_snmp baroc file for the rule base that is currently in use by the TEC Event Server or import the TeamQuest class definitions into your rule base You can add the TeamQuest class definitions to the rule base by appending the entries contained in SYS LIB TQ BASELINE TECAD BAROC to the tecad_snmp baroc file for the rule base that is currently in use by the TEC Event Server Another way to add the TeamQuest class definitions to the rule base is to import them using the TME desktop GUI or the wimprbclass command 6 The last step is to do a warm start of the SNMP Event Adapter and to stop and restart the Event Server Doing a warm start of the SNMP Event Adapter causes it to re read its configuration files These include the tecad_snmp cds and tecad_snmp oid files that were updated in steps 3 and 4 Stopping and restarting the Event Server allows it to read in the new
138. ASCII characters is the timestamp of when this execution was scheduled is the step control application group number used in this LINC system 9 bits indicates the LN or LINC format 18 bits TQ 01011 18 9 8 9 8 1 Customizing Data Collection Probes IOTRACE Data Collection Note The IOTRACE probe is released with TeamQuest Baseline only It is not available for TeamQuest Online IOTRACE is an optional feature of SIP that can run independently or with SIP data collection It gathers detailed information about every I O operation to tape or mass storage device This information is written to the IOTRACE data file SYS IODATFILE which is a tape file with standard labels or a mass storage file Information contained in each trace entry includes the file referenced device referenced run id program service times words transferred and I O function Data Collection Runstream TeamQuest Baseline contains a runstream called tqiot that processes IOTRACE data collection files The tqiot runstream uses a run identifier of tqiot and executes the tqiot program The tqiot program retrieves data from the IOTRACE file and stores it in the performance database Note The tqiot runstream is a template runstream that must be updated before it can be used You can update the template or copy the template to a different file and make updates to the tqiot processor call statement as required for your site The tqiot program pr
139. AleamQuest ClearPath OS 2200 TeamQuest Baseline and TeamQuest Online USER GUIDE Copyright 2013 TeamQuest Corporation All Rights Reserved February 2013 Release Level 7R4G TQ 01011 18 The names places and or events used in this publication are purely fictitious and are not intended to correspond to any real individual group company or event Any similarity or likeness to any real individual company or event is purely coincidental and unintentional NO WARRANTIES OF ANY NATURE ARE EXTENDED BY THE DOCUMENT Any product and related material disclosed herein are only furnished pursuant and subject to the terms and conditions of a license agreement The only warranties made remedies given and liability accepted by TeamQuest if any with respect to the products described in this document are set forth in such license agreement TeamQuest cannot accept any financial or other responsibility that may be the result of your use of the information in this document or software material including direct indirect special or consequential damages You should be very careful to ensure that the use of this information and or software material complies with the laws rules and regulations of the jurisdictions with respect to which it is used The information contained herein is subject to change without notice Revisions may be issued to advise of such changes and or additions U S Government Rights All documents product and related
140. Deletion Menu by entering 3 at the prompt When you make this selection the tqzip program prompts you one more time to verify your selection All records will then be deleted from the table Note Data deletion takes place immediately There is no opportunity to stop the deletion of records from a table Since data deletion takes place immediately it is recommended that you back up your database before you delete any records from a table Registering a Table The tqzip program allows you to register table files that you may have previously saved perhaps to tape so they can be examined with TeamQuest Baseline Because TeamQuest Baseline allows only one table to exist with a given name you will be required to rename the table you are registering A table file can contain more than one table If a file has more than one table you must specify which table from the file you want to register You are responsible for specifying the complete file name of the table file and ensuring the file is available for use when the table is accessed To register a table select Table Registration from the Tables Menu see 11 3 by entering 5 at the prompt The tqzip program displays the following Table Registration Menu gt Enter your selection lt l gt gt 5 gt gt Table Registration Menu gt gt 1 Register a Table gt 2 Deregister a Table gt gt 99 Exit Table Registration Menu TQ 01011 18 Customizing Tables Ente
141. E ACT RPT Paging Activity Summary Gives a summary of paging activity PAGE FLT RPT Paging Fault Activity Gives paging fault activity information REAL USE RPT Real Space Usage Gives real space usage for the various page states Processor Reports The processor report definitions are found in the proc subdirectory of the sip subdirectory File Name Description File Name Description File Name Description File Name Description SW QUEUE RPT Switching Queues Provides information on the switching queue when SIP level 4 or higher is configured BY UNIT RPT Proc Util by Unit Provides detailed processor utilization information for the selected processors INT SUM RPT Hardware Int Sum Summarizes hardware interrupts This information is only available if you have the hardware monitor option AVL UNIT RPT Proc Avail by Unit Provides processor availability information TQ 01011 18 Predefined Report Descriptions File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description TQ 01011 18 AVL SUM RPT Proc Avail Summar Summarizes processor availability INT UNIT RPT Processor Int by Unit Gives hardware interrupts by processor This information is only available if you have the hardware moni
142. ELINE THRESHOLDS contains the threshold definitions The following threshold definitions are included Alarm ID Description Busy IPs The IP has been fully utilized for a sustained period of time Busy Disk Disk utilization has been high for a sustained period of time Low Free Real Memory The number of free real memory pages is low Using the Default Threshold Definitions After installing and setting up TeamQuest Baseline on your system you can initiate threshold checking using the default threshold definitions Perform these steps 1 Ensure that the TeamQuest Baseline data collection probes and the tqalm program are running on your system If they are not yet running you can start them by entering start sys lib run tqctl 2 Turn on alarm checking tqzip a on 3 Load the threshold definitions into your database tqzip l e t sys lib tq baseline thresholds Note When you use the e option with the l option on tqzip the threshold definitions are loaded without first deleting any existing definitions from your database If you want to remove all existing threshold definitions from your database before loading the new ones use tqzip with the Ll option but without the e option TQ 01011 18 TeamQuest Baseline Administration 3 3 3 Changing the Default Threshold Definitions If you want to change any of the default threshold information in SYS LIB TQ BASELINE THRESHOLDS yo
143. EM System Log Data Collection Note The system log probe is released with TeamQuest Baseline only It is not available with TeamQuest Online TeamQuest Baseline contains a system log data collection runstream called tqsys The tqsys runstream uses a run identifier of tqsys and executes the tqsys program The tqsys program collects system log data and stores it in the performance database The tqsys program processes all cycles of the system log file from the previous day During the tqinit process at installation time you are prompted for the time that the tqsys runstream starts each day The tqsys program should run each day to ensure that your performance database contains a complete set of data for each day TQ 01011 18 9 33 Customizing Data Collection Probes 9 10 1 9 34 tqsys Processor Call Statement The tqsys runstream contains the tqsys processor call statement which controls the system log data reduction The tqsys processor call statement specifies data collection parameters including input parameter table name database name optional and aggregation set name Normally you use the default values for these parameters as provided in the tqsys runstream To modify any of the defaults you need to update the tqsys runstream in SYS LIB RUN The tqsys processor call has the following format tqsys z i parameter table d database name L a aggsetl END Z is used to register the tqsys probe in t
144. END keyword 9 10 TQ 01011 18 9 5 2 9 5 3 9 5 4 Customizing Data Collection Probes Terminating Data Collection To terminate the tqmips run from a demand run that has full console mode capabilities or greater enter CONS TOQMIPS END To terminate the tqmips run from the system console enter TQMIPS END Abnormal Termination of Data Collection If tgmips terminates in error due to a maximum time contingency it restarts automatically If tqmips terminates and does not restart try to resolve the problem as soon as possible and restart tqmips Data is collected only when tqmips is running Trail Selection By Date Range or Relative Time With the tqmips program you can load data from COD audit trails or from system log files into a performance database by specifying a date or relative time range This enables you to fill in gaps in the performance database with MIPS data that was collected at a time when tqmips was not running on your system When loading data using a date range tqgmips searches the system and processes all trails that contain information in the date range You must specify up to 3 additional parameters on the tqmips processor call statement See 9 5 1 for the format of the processor call statement The format for the date range parameters is b mmddyy yy e mmddyyLyy n audit trail number where b specifies the month day and year of the beginning date used in the log file selecti
145. Group Derived Category Derived Category Subcategory Derived Subcategory Statistic lt your statistic name gt TQ 01011 18 8 1 Customizing Derived Statistics 8 2 2 Parameter References and Substitution Derived statistic formulas can reference the values of other derived statistics or of parameters being collected by the TeamQuest Baseline data collection probes This allows you to calculate new statistics based on existing ones Parameter references are made by specifying all of the key names used to identify a parameter in the database See 2 2 2 for a description of the parameter key name hierarchy To keep the formula definition itself as short and concise as possible you define short names called aliases for each of the parameters you wish to reference from within the formula The list of aliases used in your formula is specified when you create your derived statistic using TeamQuest Performance Software Administration Parameter references in derived statistic formulas can be either specific references that contain the entire set of key names needed to identify the parameter or they can contain substitution indicators Substitution indicators are used to tell TeamQuest Baseline to fill in the values for selected keys at report time Substitution is allowed for system workload set workload resource and application group The following special strings are used in a parameter reference to indicate that substit
146. Group Category Subcategory Statistic Name Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec Deadlocks Sec TQ 01011 18 IO Activity Multihost Summary Total lock requests per second Total unlock requests per second Total delayed response messages per second Total deadlocks per second Total lock requests Total unlock requests Total delayed response messages Total deadlocks IO Activity Multihost TIP Total lock requests per second by TIP Total unlock requests per second by TIP Total delayed response messages per second by TIP Total deadlocks per second by TIP Total lock requests by TIP Total unlock requests by TIP Total delayed response messages by TIP Total deadlocks encountered by TIP IO Activity Multihost UDS Total lock requests per second by UDS Total unlock requests per second by UDS Total delayed response messages per second by UDS Total deadlocks per second by UDS A 33 OS 2200 Parameters A 34 Lock Reqs Unlock Reqs Delayed Resp Msg Deadlock Total lock requests by UDS Total unlock requests by UDS Total delayed response messages by UDS Total deadlocks encountered by UDS The following statistics are created for UDS when you have an XPC and the UDSTRP SIP counter Parameter Hierarchy Category Group Category Subcategory Statistic Name Lock Rec Lock Blk Lock File Maint Lock Immed Lock Rec Immed Lock Blk Lock String Lock Blk Rec Prb Lock Rec Prb Lock R
147. IO SUPs Core Block SUPs IO Requests Disk I O Requests Disk I O Words Tape I O Requests Tape I O Words Voluntary Wait Resource Wait Active Time Cards In Runs Opened Max Real Memory Real Memory Usage A 50 System Log Workload Deadline Total SUPs used to process the job Sum of the IP SUPs ER CC SUPs operating system SUPs and I O SUPs Total IP SUPs Total ER and control card SUPs Total Exec Services I O SUPs Total Exec Services IP SUPs Total control statement I O SUPs Total control statement IP SUPs Total I O SUPs Total core block SUPs Sum of SUPs and main storage usage Total number of I O requests made by this run Total number of I O requests to disk that have been made by this run Total number of words that have been transferred during disk I O operations by this run Total number of I O requests to tape that have been made by this run Total number of words that have been transferred during tape I O operations by this run Total voluntary wait time Total amount of time the run spent waiting for a resource Total time the job was active Total number of cards the Exec read during the run Total number of runs that were opened Maximum number of real pages concurrently resident for this run The product of the total number of resident pages and the time in seconds that the pages are resident TQ 01011 18 OS 2200 Parameters A 8 Workload Performance Derived Statistics Team Quest Bas
148. IP Txns Sec TIP Txns Sec MIN TIP Txns Sec MAX TIP Txns Completed Sec Demand Resps Sec Batch Runs Opened Sec Demand Runs Opened Batch Runs Opened Total Runs Opened Category Group Category Subcategory Statistic Name Reqs Sec Category Group Category Subcategory Statistic Name Reqs Sec Control statement SUP charges by non TIP programs in seconds Control statement SUP charges by TIP transactions and batch connect programs in seconds IP SUP charges by non TIP instruction processor execution in seconds IP SUP charges by TIP instruction processor execution for TIP transactions and batch connect programs in seconds User I O SUP charges to non TIP files in seconds User I O SUP charges to TIP files in seconds System Throughput Number of TIP transactions scheduled per second Minimum number of TIP transactions scheduled per second Maximum number of TIP transactions scheduled per second Number of TIP transactions completed per second Number of demand responses per second Number of batch runs opened per second Total demand runs opened Total batch runs opened Total of newly opened runs System Exec Requests System Total Requests per second for all Executive Requests ER System Exec CALLs System Total Requests per second for all call gates defined on the system TQ 01011 18 OS 2200 Parameters A 2 2 Category Group Category Resource Subcategor
149. If the category group is not specified User Defined is used for the category group c category specifies the category key name in the parameter hierarchy of the statistic being stored If the category is not specified null is used for category b subcat specifies the subcategory key name in the parameter hierarchy of the statistic being stored If the subcategory is not specified null is used for subcategory 9 40 TQ 01011 18 Customizing Data Collection Probes a aggset1 specifies the aggregation set to use for storing data The aggset1 parameter is the name of an aggregation set defined in the database An aggregation set is strictly optional on the command line If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used h prints usage information containing command line syntax and exits END required keyword which ends command line option scanning TQARG File Input Format The TQARG file consists of a line of statistics followed by a line of keywords describing the statistic s weights The following example shows the required format of the TQARG file used by the tqusrprb program DATE TIME total used available used RESOURCE NONE NONE LAST LAST LAST LAST NONE w gt The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The first line of input must co
150. Internet RFCs TQ2200 MIB DEFINITIONS BEGIN IMPORTS Enterprises FROM RFC1155 SMI OBJECT TYPE FROM RFC 1212 DisplayString FROM RFC1213 MIB TRAP TYPE FROM RFC 1215 aH TeamQuest enterprise tq2200 OBJECT IDENTIFIER enterprises unisys 223 os2200 7 6 i TeamQuest enterprise groups The only group defined initially i is the TeamQuest Alarm Trap group Other groups will be added as a4 required tq2200AlarmTrap OBJECT IDENTIFIER tq2200 1 The tqAlarmTrap group This group contains objects describing ae the TeamQuest SNMP Alarm Trap TQ 01011 18 J 9 2200 SNMP Trap Alarms tq2200Alar SYNTAX ACCESS STATUS DESCRIP The al on whi tq2200Alar SYNTAX ACCESS STATUS DESCRIP The al parame qual tq2200Alar SYNTAX norma ifi System OBJECT TYPE DisplayString SIZE 0 read only andatory TION arm system is a textual ch the alarm was raised tq2200AlarmTrap 1 52 string that identifies the system Identifier OBJECT TYPE DisplayString SIZE 0 read only andatory TION arm identifier is a textual string that identifies the ter for this alarm The qualifierl qualifier2 and er3 fields are appended to this field if present tq2200AlarmTrap 2 244 Severity OBJECT TYPE NTEGER 1 warning 2 critical 3 ACCESS read only STATUS mandatory DESCRIPTION The severity of this alarm as determined by the change in zones tq2200AlarmTra
151. MINUTE HOUR DAY WEEK MONTH or YEAR A value is required for this field and it must be greater than 0 RetentionSpec The interval and the length of time the records created by the data collectors are retained Zero or more lines of RetentionSpec are allowed Each retention specification is on a separate line and is made up of two values separated by a space or a comma RetentionSpec interval retention period The interval is specified as hhmmss The retention period is specified as nUNITS n must be an integer greater than zero and UNITS can be MINUTE HOUR DAY WEEK MONTH or YEAR A value is required for both interval and retention period in each RetentionSpec TQ 01011 18 11 23 Customizing Tables Collection CollectionSpec 11 24 The set of RetentionSpec lines in the load file replaces any RetentionSpec lines previously defined for the table If no RetentionSpec lines are specified in the load file all existing RetentionSpec lines are deleted for the table and no new ones are added RetentionSpec lines are ignored if the table does not contain intervalized data Indicates whether or not data collection is turned on for the table If data collection is to be done for the table specify ON Otherwise specify OFF If no value is specified ON is used as the default The start and stop time the interval and the exclusion mask for data collection Zero or more lines of CollectionSpec are allowed Each collectio
152. OS 2200 Parameters A 2 6 SIP Parameters for the XPC The following is a list of the statistics collected by the SIP probe for the XPC Parameter Hierarchy Category Group IO Activity Category XPC Resource XPCO XPC1 Subcategory Summary Statistic Name Reqs Sec XPC requests per second Words Req XPC words per request Avg RET XPC average request existence time ms Function Rsp Time XPC average response time for all XPC functions ms Exec Staged XPC number of words per second staged by the Exec Words Sec Exec Destage XPC number of words per second destaged by the Exec Words Sec Discarded Dst Reqs S XPC initiated destage requests per second that were discarded Destage Avoided XPC destage requests per second that were avoided Regqs S Disk Traffic Percent of disk traffic for XPC requests in comparison to a disk only environment Hit Rate XPC hit rate as a percentage XIIP Queue Length Average initiation queue length of all XIIPs connected to this XPC Programs Sec Average XPC programs per second executed by the XPC Program Rsp Time Average XPC response time in milliseconds for an XPC program XPC L UD Synch Average number of XPC L UD Synch entry write requests Reqs Sec per second A 36 TQ 01011 18 OS 2200 Parameters Category Group Category Resource Subcategory Statistic Name IXP Available Written to File Storage Category Group Category Resource Subcategory Statistic Name Reqs
153. Other users can view derived statistic information in read only mode when the database is locked TeamQuest Performance Software Administration allows you to perform the following tasks e Performing Derived Statistic Administration Tasks see 8 3 1 e Creating a New Derived Statistic Definition see 8 3 3 e Saving a Derived Statistic Definition see 8 3 4 e Viewing Long Key Names see 8 3 5 e Importing Derived Statistic Definitions see 8 3 6 e Exporting Derived Statistic Definitions see 8 3 7 e Displaying or Modifying a Derived Statistic Definition see 8 3 8 e Deleting a Derived Statistic Definition see 8 3 9 Note Itis recommended that you use the Administration utility to create and update your derived statistic definitions TQ 01011 18 Customizing Derived Statistics 8 3 1 Performing Derived Statistic Administration Tasks To begin any derived statistic maintenance task select Administration gt Derived Stats If more than one session is open the Administration Session Selection dialog box is displayed This dialog box contains a list of open sessions for which you can perform derived statistic maintenance and session information for the session highlighted in the list box Select the desired session and click OK If more than one database is available for the session the Administration Database Selection dialog box is displayed Select the database for which you want to perform derived statistic maintenanc
154. P The following statistics are collected by the SIP probe for the 2200 Queueing Architecture Parameter Hierarchy Category Group Category Subcategory Statistic Name Enqueued Sec Queued Time msec Category Group Category Subcategory Statistics Name Activations Sec Category Group Category Subcategory Statistic Name Exceeded Sec TQ 01011 18 Queue System Return Entries Number of queue entries enqueued per second to the system return queue Time in milliseconds per entry spent on the system return queue Queue System Return Server Rate at which the system return queue server is activated Queue System Return MaxCount Number of MaxCount exceeded interrupts per second for the system return queue A 39 OS 2200 Parameters A 40 Category Group Category Subcategory Statistic Name Enqueued Sec Queued Time msec Category Group Category Subcategory Statistic Name Expansions Sec Contractions Sec Category Group Category Statistic Name Deactivations Sec Queues Category Group Statistic Name Banks Queue Available Entries Rate at which entries are enqueued to the available queue Time in milliseconds per entry spent on any available queue Queue Available MaxCount Number of MaxCount expansions per second for any available queue Number of MaxCount contractions per second for any available queue Queue Available Rate at which
155. PC RET RPT XPC Request Existence Time Provides request existence time information for the selected XPC broken down by XPC function XPC REQ RPT XPC Requests Second Provides requests per second information for the selected XPC broken down by XPC function XITP ACT RPT Activity by XIIP Gives specific XIIP performance information Java Reports The java report definitions are found in the java subdirectory of the sip subdirectory Statistics in the following reports are available on systems where it is possible to designate a processor or set of processors as dedicated Java IPs File Name Description File Name Description File Name Description File Name Java Description BY UNIT RPT Java Proc Util by Unit Provides detailed Java processor utilization information for the selected processors INT SUM RPT Java Hardware Int Sum Summarizes Java hardware interrupts This information is only available if you have the hardware monitor option AVL UNIT RPT Java Proc Avail by Unit Provides Java processor availability information AVL SUM RPT Proc Avail Summar Summarizes Java processor availability TQ 01011 18 Predefined Report Descriptions B 3 3 File Name Description File Name Description File Name Description File Name Description INT UNIT RPT Java Processor Int by Unit Gives hardware interrupts by Java processor Thi
156. PRNT1 Summary Device percent busy Number of requests per second Percentage of all requests that were reads Percentage of all requests that were queued Average number of K words per request Average number of K words per request that were input Average number of K words per request that were output Average transfer time per request Average service time Average hardware service time Concurrency IO Activity Cache Disk Cache disk destage requests per second Cache disk destage avoided Cache disk speculation success Cache disk traffic Cache disk hit rate A 29 OS 2200 Parameters A 30 Age Oldest Segment Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Cache disk age of the oldest segment Cache disk request per second Cache disk K words per request Cache disk transfer time Cache disk service time Cache disk hardware service time IO Activity Cache Disk Read Hit Cache disk read hit requests per second Cache disk read hit K words per request Cache disk
157. Parameters Appendix B TQ 01011 18 A 1 A 2 A 2 1 A 2 2 A 2 2 1 A 2 2 2 A 2 3 A 2 3 1 A 2 4 A 2 5 A 2 6 A 10 A 10 1 A 10 2 A 10 3 A 10 4 Introduction SIP Parameters System Statistics Processor Statistics Performance Monitor Dedicated Java Processor Statistics Memory Statistics Program Load Statistics Paging Statistics I O Activity Statistics SIP Parameters for the XPC Queue Statistics TPM Log Parameters MIPS Parameters MIPS Probe Derived Statistics IOTRACE Parameters MAPPER Log Parameters System Log Parameters Workload Performance Derived Statistics SIP Probe Derived Statistics MAPPER Log Probe Derived Statistics System Log Probe Derived Statistics System Activity Monitor Statistics Runs Table Transaction Table Open Distributed Transaction Processing Statistics System Table Information Server Table Information Client Table Information Service Table Information Predefined Report Descriptions B 1 B 2 B 3 B 3 1 B 3 2 B 3 3 B 3 4 B 3 5 B 3 6 Introduction Report Definition Directory SIP Reports I O Reports Java Reports Memory Reports Processor Reports System Reports Summary Reports IOTRACE Reports MIPS Report MAPPER Log Reports System Log Reports TPM Log Reports Workload Performance Reports A 1 A 2 A 5 A 7 A 9 A 14 A 18 A 18 A 22 A 24 A 36 A 39 A 41 A 43 A 43 A 44 A 45 A 46 A 51 A 51 A 53 A 54 A 58 A 58 A 59 A 61 A 61 A 62 A 64 A 64
158. Positive integer StatisticItem LastPoint Bottom Positive integer StatisticItem LastPoint The following descriptions explain the syntax of the parameter group specifications of the report definition parameter group The name of the data series This is a unique name by which the data series is known throughout the report definition The parameter group can be a string of up to 24 alphanumeric characters Keyname keyvalue keyvalue The definitions for the host defined keywords The host defines the keywords for the Keyname specification If a keyword used for a Keyname specification is different from the one defined by the host the specification is ignored Some Keyname specifications can contain multiple keyvalue specifications If you supply multiple keyvalues for a Keyname specification that does not allow multiple values all values but the first value are ignored Keyname For the Team Quest Performance Software the Keyname is one of the following keywords defined by the host system Statistic related key names Category Group Category Subcategory Statistic Note On some platforms Category and Subcategory are combined into the form Category Subcategory and called Class and Subclass in the TeamQuest View interface TQ 01011 18 C 29 Report Definition File Qualifier related key names System Workload Set Workload Resource Application Group Unisys OS 2200 only keyvalue C
159. Q 01011 18 Customizing Alarms and Thresholds The alarm identifier in each TeamQuest Baseline alarm message should be used within SMART SP AMS pattern matching to determine which alert needs to be raised or cleared This alarm identifier can be used as the alarm identifier in the alarm event report SMART SP AMS creates and sends to SPO If you include the appropriate user text in the alert event report it will be displayed by SPO in the Alerts and Alert Detail windows In addition you can define supplemental help text for the alarm and keep this file on the SPO server It will be displayed in the Alert Detail window SPO lets you define alarm policies that include taking actions when alarms are raised These actions include dialing a pager both immediately and after a timed delay This permits you to notify a site support person when TeamQuest Baseline discovers a performance anomaly Then if the problem has not been corrected after a certain amount of time you can automatically escalate the notification The TeamQuest Baseline alarm message contains an indicator of the alarm severity The indicators are NORMAL WARNING and CRITICAL The alarm severities can be used within SMART SP AMS pattern matching to determine when to raise or clear an alert A severity of WARNING or CRITICAL indicates that an alert should be raised A severity of NORMAL indicates that the alert can be cleared Single Point Status provides enhanced monitoring of system
160. Quest Baseline data collection probes are initiated on the host system These probes start the data collection and automatically process the data They also extract statistical values and date and time information for performance parameters and store that information in the performance database TeamQuest Baseline and TeamQuest Online come packaged with the tqbsp probe The tqbsp probe collects Software Instrumentation Package SIP data from the Exec Using the ER SYSBAL mechanism the tqbsp probe starts the data collection and automatically processes that data The tqbsp probe extracts statistical values for performance parameters and stores this information in the performance database TQ 01011 18 Introduction TeamQuest Baseline comes packaged with the following additional probes that provide the workload capability of the product TQ 01011 18 tqsam The tqsam probe collects information on active runs and transactions using calls to Exec INSP TeamQuest Baseline contributes additional capability to dynamically put the data collected by the tqsam probe into workloads and to change various collection attributes tqrtpm The tqrtpm probe processes TIP HVTIP statistics from the TIP Performance Monitor TPM audit trails created when TIP logging is configured This probe requires that your TPM audit trail be configured to mass storage or moved to mass storage prior to starting either probe The tqrtpm probe runs continuously to report
161. SIP is turned off at the console tqbsp turns SIP back on automatically and uses the ER SYSBAL interface upon restarting to continue data collection If tqbsp terminates and does not restart try to resolve the problem as soon as possible and restart tqbsp Data is collected only when tqbsp is running TQ 01011 18 9 3 Customizing Data Collection Probes 9 3 4 SIP File Input With the tqbsp program you can load SIP data directly from a SIP file into a performance database This method enables you to load SIP data from another system or to update the performance database with SIP data that was generated at a time when tqbsp was not running on your system To load SIP data directly from SIP files you must specify three additional parameters on the tqbsp processor call statement See 9 3 1 for the format of the processor call statement The format for these parameters is as follows a aggsetl f sipfile n l f sipfile ntl S system name where a specifies the aggregation set to be used for data collection The aggset1 parameter is the name of the aggregation set defined in the database name The aggset1 parameter is case sensitive Only one aggregation set is allowed f specifies a SIP file to be processed sipfile is the name of the SIP file n is the file cycle number When processing multiple system log files multiple f options are needed one for each cycle to be processed All cycles must be in sequential orde
162. Sec Words Req Avg RET Response Time Category Group Category Resource Subcategory Statistic Name Reqs Sec Words Req Avg RET Response Time Category Group Category Resource Subcategory Statistic Name Reqs Sec Words Req Avg RET Response Time TQ 01011 18 IO Activity XPC XPCO XPC1 Utilization Average percentage of time that each IXP is busy Percentage of time the XPC Performance Monitor is available Average percentage of XPC file acceleration storage that has been written to file storage IO Activity XPC XPCO XPC1 Read Hit XPC read hit requests per second XPC read hit words per request XPC read hit average request existence time XPC read hit average response time IO Activity XPC XPCO XPC1 Write Hit XPC write hit requests per second XPC write hit words per request XPC write hit average request existence time XPC write hit average response time IO Activity XPC XPCO XPC1 Read Miss XPC read miss requests per second XPC read miss words per request XPC read miss average request existence time XPC read miss average response time A 37 OS 2200 Parameters A 38 Category Group Category Resource Subcategory Statistic Name Reqs Sec Words Req Avg RET Response Time Category Group Category Resource Subcategory Statistic Name Reqs Sec Words Req Avg RET Response Time Category Group Category Resource Subcategory St
163. Software Administration allows you to perform the following tasks e Performing Aggregation Set Administration Tasks see 6 3 1 e Creating an Aggregation Set see 6 3 2 e Displaying or Modifying Aggregation Set Definitions see 6 3 3 e Deleting an Aggregation Set see 6 3 4 Note It is recommended that you use TeamQuest Performance Software Administration to maintain your aggregation sets After you create remove or change an aggregation set you must customize your data collection to take advantage of any aggregation changes see 6 4 7 TQ 01011 18 6 3 1 Customizing Aggregation Sets Performing Aggregation Set Administration Tasks To begin any aggregation set maintenance task select Administration gt Aggregations If more than one session is open the Administration Session Selection dialog box is displayed This dialog box contains a list of open sessions for which you can perform aggregation set maintenance and session information for the session highlighted in the list box Select the desired session and click OK If more than one database is available for the session the Administration Database Selection dialog box is displayed Select the database for which you want to perform aggregation set maintenance and click OK The Aggregation Set Maintenance dialog box is displayed Aggregation Set Maintenance Session cl01 Database PRODUCTION Aggregation Sets Delete All aggregation set main
164. String Qualifier_2 String Qualifier_3 String TQ 01011 18 Description The date and time at which the alarm occurred A sequential number assigned to the alarm for uniqueness The name of the system for which the alarm occurred The severity of the alarm NORMAL WARNING or CRITICAL The user defined alarm identifier assigned to this alarm Either the resource name for example diskname or application group name for which the alarm was generated If both resource name and application group name apply to the statistic Qualifier_1 contains both names separated by a space The workload set name for which the alarm was generated The workload name for which the alarm was generated Alarm History Table Statistic Parameter_Value User_Text_1 User_Text_2 User_Text_3 First_Threshold Second_Threshold Count Parameter_Name Database_Name String Real Number String String String Real Number Real Number Integer String String The name of the statistic including the category group category subcategory and statistic name The parameter value that caused the alarm to be generated The user text line 1 from the threshold definition for the alarm The user text line 2 from the threshold definition for the alarm The user text line 3 from the threshold definition for the alarm The first threshold from the threshold definition for the alarm The second threshold from the thres
165. TQZIP gt gt gt Enter your selection gt 2 The tqzip program then displays the Tables Menu Tables Menu gt gt gt 1 List Tables gt 2 Collection Periods gt 3 Retention Periods gt 4 Record Deletion gt 5 Table Registration gt gt gt gt 99 Exit Tables Menu Enter your selection lt 1 gt To perform a maintenance task enter its associated number at the prompt To return to the Administration Menu enter 99 When you select the various maintenance tasks for the Tables Menu the tqzip program prompts you to supply values or other information Default values are displayed for each prompt in angle brackets lt gt To use a default press the transmit key without entering a value TQ 01011 18 11 3 Customizing Tables 11 4 11 5 11 4 Displaying a List of Tables You can view a list of the tables in your database that contain performance information To display the list select List Tables from the Tables Menu see 11 3 by entering 1 at the prompt gt Enter your selection lt 1 gt gt 1 gt gt Class Name Rev File gt a Se aan ci ee peck pia Say Eas ees an eect Re en fide be hot ates eo aa iah heed oa oS eae fet tose tee gt 2200 Runs 1 tqdata prodsamfie gt Alarm AlarmLog 2 tqdata prodalarmlg gt 2200 Transactions 1 tqdata prodsamfie gt gt Continue The list of tables generated by tqzip includes the table class table name
166. TRACE and MAPPER systems tqagg tqalm tqzip tqrep and tqrpt TQ 01011 18 Introduction Data Collection Exec HVTIP MAPPER IOTRACE System Probes tqagg tqalm a Database Management Performance Database A A tqzip A Reports y Reports tqdata gt tqrpt gt W ae Reports y an Printer Files lt TeamQuest a a View oe Exported Report Data Display Data Files Figure 1 2 TeamQuest Baseline Components TQ 01011 18 Introduction 1 6 1 6 1 1 6 1 1 How Do TeamQuest Baseline and TeamQuest Online Work TeamQuest Baseline and TeamQuest Online consist of a host software component and the TeamQuest View user interface software The host software collects performance data stores this data in the performance database and maintains the database TeamQuest View requests performance data from the performance database and creates reports The Host Software Component The TeamQuest Baseline and TeamQuest Online host software runs on your OS 2200 system collecting data and maintaining the performance database The host software uses probes to collect data Individual probes collect data from different sources Once the data is collected TeamQuest Baseline or TeamQuest Online stores the data for future retrieval Probes The Team
167. TeamQuest View User Guide for information on User Preferences Time Format format The specifications for the format of the time as it appears on the report format Can be any of the following and must be entered in lowercase hh Hours 00 23 hhh Hours 00 12 mm Minutes 00 59 ss Seconds 00 59 pp AM or PM indicator The default values for Time are defined in the Date Time Formats settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Aggregation Set aggregation set name The name for the aggregation set in the database to use for data retrieval Report Type Monitor History The type of report for the report definition You can specify either Monitor or History Monitor Indicates that this report is used for real time monitoring History Indicates that this report is used for reporting historical data If this field is not present the report defaults to the calendar from the session Date Multiplier positive integer The number used to consolidate data across time groups in the aggregation set For example if daily data is stored in the database and a date multiplier of 7 is specified the report shows data consolidated into one sample for every seven days of data The specification applies to history reports only positive integer Is a positive integer that is used as the date multiplier The default value is 1 A value less than 2 causes no consolidati
168. UPs Total ER and control card SUPs Total Exec Services I O SUPs Total Exec Services IP SUPs Total control statement I O SUPs Total control statement IP SUPs Total I O SUPs Total core block SUPs Sum of SUPs and main storage usage Total number of I O requests made by this transaction Total number of I O requests to disk that have been made by this transaction Total number of words that have been transferred during disk I O operations by this transaction Total number of I O requests to tape that have been made by this transaction Total number of words that have been transferred during tape I O operations by this transaction Total voluntary wait time Sum of user response time and TWAIT time Total amount of time the transaction spent waiting for a resource Total time the transaction was active Total number of transactions Maximum number of real pages concurrently resident for this transaction The product of the total number of resident pages and the time in seconds that the pages are resident A 47 OS 2200 Parameters Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs IO SUPs ES IO SUPs ES IP SUPs CS IO SUPs CS IP SUPs Core Block SUPs IO Requests Disk I O Requests Disk I O Words Tape I O Requests Tape I O Words Voluntary Wait Resource Wait Active Time Cards In Runs Opened Max Real Memory Real Memory Usage A 48 System Log Workload Dem
169. When a connection is made TQTCP passes the Team Quest View or TeamQuest Performance Software Administration conversation to a tqdata session that is executing in a new RSI run Both the TQTCP executable and background runs are installed with TeamQuest Baseline You must also update the configuration element in SYS LIB TQ BASELINE TQTCP CONFIG before starting the TQTCP background run The TQTCP background run must be active on your system whenever you are using a socket connection to your 2200 system through TeamQuest View or TeamQuest Performance Software Administration TQ 01011 18 2200 Socket Interface 1 2 1 Background Run The TQTCP runstream is installed into SYS LIB RUN You can start the TQTCP background run by entering the following keyin at the system console ST TQTCP To start the background run from a demand or batch run use the following statement start sys lib run tqtcp By default the TQTCP runstream executes TQTCP without processor call options If you need to use any options you will need to update the TQTCP runstream accordingly before it is started The background run provides a keyin interface for connection status reporting and TQTCP termination By default the keyin name is TQTCP You can use the k option on the TQTCP processor call to change this keyin name if needed A keyin is executed by entering the keyin name followed by one of these keyin values STATUS provides information on open connecti
170. XX XXX XXX XXX For more information on how to update your CMS 1100 configuration for using SNMP traps see the OS 2200 Communications Management System CMS 1100 Installation and Configuration Guide For information on the PROCESS and SNMP MGMT statements see the OS 2200 Communications Management System CMS 1100 Configuration Reference Note The Unisys product CPComm also supports the SNMP TSAM interface required for sending SNMP traps If you are using CPComm you will need to make the appropriate configuration changes for defining the TeamQuest SSU and trap destinations TeamQuest Alarm Administration Interface This feature is implemented as part of the Alarms and Thresholds capability of the TeamQuest Performance Software see Section 7 When a threshold is defined using the tqzip program you can specify an SNMP trap action by placing a special syntax in one of the user text fields of the threshold definition When an alarm is raised and this special syntax is present an SNMP trap is sent according to the specifications in the syntax Specifically an SNMPv1 trap is sent with a generic trap type of enterprise specific 6 and a specific trap type that is used to further identify the contents of the trap see J 4 1 for a list of the specific trap types Information about the alarm is placed in the variable bindings section of the trap and is available for display on a management console or any other use There are 1
171. a that is kept as it is stored by the data source and is called the primary retention For tables that contain intervalized data additional retention specifications can be defined These subsequent retention specifications follow the primary retention and define how long and at what interval data is retained It is assumed that larger intervals follow smaller intervals For example you might define a 5 minute 2 week retention and a 30 minute 3 month retention The data is kept at 5 minute intervals for two weeks and then an additional three months at 30 minute intervals This results in a total of three months and two weeks worth of data being retained TQ 01011 18 11 9 Customizing Tables The data in the primary retention specification is placed in the table by data collectors The data in subsequent retentions is created by the TeamQuest Baseline aggregator based on the originally collected data When the end of a retention period has been reached the Baseline aggregator removes the data from the table When the aggregator creates new records it will synchronize them with even intervals for a day You supply this retention start time when you create the retention specifications Normally a retention start time of midnight is adequate Thus for 30 minute data points you would have records created at 00 00 00 30 1 00 1 30 2 00 and so on There are times when you may want the synchronization time to be something other than midnight For
172. ach workload on the total IP ER CC and I O SUPs TXNCNT RPT Transaction count Gives the statistics for each workload on the number of transactions number of passoff messages number of I O requests and response processing and termination average commit time RTPS RPT RTPS counts Gives statistics for each workload on the number of RTPs deactivated transactions and the number of RTPs deactivated copies with the same transaction code TXNACT RPT Transactions active Gives statistics for each workload on the number of stuck copies total stuck active copies and total active counts TQ 01011 18 Predefined Report Descriptions B 9 Workload Performance Reports The workload performance reports provide information about workload service levels and facility utilizations These reports are found in the wkldperf subdirectory and provide derived statistics information from the SIP tqbsp TPM log tqtpm MAPPER log tqmpr and system log tqsys probes These reports are intended for top down performance analysis activity that may include capacity planning Note Ifyou open these reports and no data values are reported for any of the statistics you may need to load the workload performance derived statistic definitions into your database see A 8 File Name Description File Name Description File Name Description File Name Description TQ 01011 18 summary rpt Service Leve
173. active TeamQuest Baseline probes are automatically notified of the threshold changes They begin using the new threshold information with the next data sample they collect To load threshold definitions with tqzip use the following processor call format tqzip d database 1 t threshold file e where d specifies the database that tqzip uses The database parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 l specifies that a load operation is to be performed by tqzip The type of load operation is thresholds as indicated by the t option t loads the threshold definitions into the database from the files specified by the threshold file parameter The threshold file parameter is the name of a threshold definition file created by the tqzip u and t options see 7 13 e loads the new threshold definitions from the threshold definition file without first deleting the existing definitions in the database If you do not specify the e option all existing definitions are deleted from your database before the new ones are loaded Example By using the l and t options of the tqzip processor call you can use different sets of threshold definitions at various times of the day For example you may want to use different threshold values for your day shift than you use for the night shift You can do this b
174. ad you define will be inserted after the selected item and belong to the selected workload set Notice that the term workload set appears under the New button indicating that you are going to create a workload definition in the database 4 Workload Maintenance Session cl01 Database PRODUCTION Workload Info Workload information received in Update mode The order that workloads are defined is very important When a probe processes an input record the workloads are evaluated in the order specified when they are defined Once a probe determines that the data record belongs to a workload the data is assigned to only that workload The remaining workloads are not evaluated If the record does not belong to any defined workload it is assigned to a default workload named OTHER The sum of all workloads including OTHER represents all the work performed on the system Note You can use the order of your workload definitions to streamline the evaluation process If workloads that catch the most records are placed toward the top of the workload definition list fewer evaluation steps are required and the evaluation process is completed more quickly Customizing Workloads When you click New the Workload Definition dialog box is displayed 4 Workload Definition Session cl01 Database PRODUCTION Source Data Type hvslog Workload Set Name primary Cancel Workload Name hs ttst CS Help Workload Characterization C
175. ajorTick field is crossing MinorTick Defines the type of marks used for the minor ticks on the vertical axis line The acceptable values are Invisible No ticks marks appear Inside Tick marks appear on the inside of the axis line Outside Tick marks appear on the outside of the axis line Crossing Tick marks cross the axis line The default value for the MinorTick field is invisible C 12 TQ 01011 18 Report Definition File MajorGrid Indicates whether grid lines are drawn across the graph at the points of the vertical axis designated by the major ticks The acceptable values are ON or OFF If MajorGrid is set to ON the grid lines appear on the report If MajorGrid is set to OFF the grid lines do not appear on the report The default value is OFF MinorGrid Indicates whether grid lines are drawn across the graph at the points of the vertical axis designated by the minor ticks The acceptable values are ON or OFF If MinorGrid is set to ON the grid lines appear on the report If MinorGrid is set to OFF the grid lines do not appear on the report The default value is OFF The default values for Value Line are defined in the Axis Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Values Major Minor Minimum Maximum Metric The specifications for the presentation of parameter values on the report The Values specifications determine the minimum and maximum va
176. alarm occurred The symbolic host name is also available in the System field of the variable fields see J 4 2 Generic Trap This field is set to the enterprise specific 6 trap type in all cases Specific Trap This value is used to further identify the contents of the trap 1 TeamQuest Performance Software alarm with a severity of Normal or Harmless 2 TeamQuest Performance Software alarm with a severity of Warning 3 TeamQuest Performance Software alarm with a severity of Critical Timestamp This field is set to sysUpTime which is the amount of time since the host was started TQ 01011 18 2200 SNMP Trap Alarms J 4 2 Variable Fields The following fields are specific to the TeamQuest Baseline SNMP trap and are defined formally in the TeamQuest Enterprise Management Information Base MIB See J 6 for more information on the MIB These variable fields are mapped as follows System This field is set to the name of the host that generated the alarm Identifier This field is set to the Alarm Identifier value from the threshold definition and should contain a description of the entity or variable that raises the alarm Severity Severity is based on the difference between the current and normal zone values In all situations the severity is NORMAL if the current and normal zone match In situations where 2 zones are defined the zone difference will be either 0 or 1 If the difference is 1 the severity is CRITICAL
177. alert box the workload name is removed from the Workload Info list box in the Workload Maintenance dialog box To delete the workload definition from the database you must click OK in the Workload Maintenance dialog box You can cancel the deletion of a workload either by clicking the No button in the alert box or by clicking Cancel in the Workload Maintenance dialog box see 5 3 1 Note By deleting a workload definition from the database neither any database parameters are deleted nor are any data values cleared This is because deleting individual workload parameters and data values would cause the data previously stored for that workload set to be incomplete Maintaining Workloads with tqzip The tqzip program allows you to perform the following additional tasks associated with maintaining the workload capability e Unload workload definitions contained in the database to a file see 5 4 1 You can use this file to view the definitions and update the definitions by using a text editor You can then reload the definitions e Load workload definitions from a specified file into the database see 5 4 2 TQ 01011 18 5 4 1 Customizing Workloads Unloading Workload Definitions The tqzip program allows you to export all the workload definitions currently contained in the performance database into a specified file You can use this file to view the workload definitions and update the workload definitions by using a text editor
178. alid field names for the table The data will be sorted by the specified field in either ascending or descending order If the sorted order results in duplicate sort keys the second field if specified will be used to sort these records and so on Filter text The criteria to be used to filter records for display Only records that match the criteria defined in the text filter statement The filter statement must be enclosed in double quotation marks See 5 2 2 for a description of the definition language that can be used to create filter statements Combine field1 field2 The fields to use for combining data records field 1 field2 can be any valid field name for the table The records will be combined where the values of the given field name match exactly Using Special Identifiers in Section Keys Special identifiers can be used as key values in both the Report section and the Table section of the report definition file The special identifiers allow for substitution of values at the time the report is generated By using special identifiers you do not need to supply actual values in the report definition A special identifier consists of an ampersand amp followed by a valid identifier name If an ampersand is followed by an unknown identifier the ampersand and identifier are printed in the report as they appear in the report definition Special identifiers are not case sensitive The following are valid spec
179. all statement 9 20 workload characteristics 9 23 tqunlk program description 10 7 processor call statement 10 7 tqusrprb probe data storage 9 45 description 1 10 9 39 installing 9 47 PRBDATA file input format 9 39 processor call format 9 39 TQARG file input format 9 39 writing a probe for performance data 9 39 tqzip program adding collection period 11 7 retention specifications 11 13 Alarms and Thresholds Menu 7 5 applying collection period changes 11 9 retention specification changes 11 15 TQ 01011 18 Index Collection Period Menu 11 4 creating derived statistic threshold definitions 7 14 threshold definitions 7 6 deleting collection periods 11 8 records 11 16 retention specifications 11 14 threshold definitions 7 19 deregistering a table 11 20 displaying a list of tables 11 4 exiting Collection Period Menu 11 9 Retention Specification Menu 11 15 listing collection periods 11 6 parameters 10 4 retention specifications 11 11 loading derived statistic definitions 8 21 threshold definitions 7 24 workload definitions 5 22 maintaining alarms 7 1 thresholds 7 1 modifying primary retention specification 11 12 notifying probes of threshold changes 7 21 processor call format H 1 registering tables 11 18 Retention Specification Menu 11 9 reusing threshold definitions 7 11 setting retention start time 11 15 threshold checking 7 21 turning on and off data collection 11 9 threshold checking 7 25 unloading d
180. allows you to select another threshold for deletion When any changes have been made to the threshold definitions in the performance database the tqzip program displays a final prompt If overall threshold checking is turned on you must notify any active probes of the threshold definition changes before they will take effect Enter y at the prompt to notify the probes You can also use the Notify Probes of Threshold Changes selection on the Alarms and Thresholds Menu to perform this task see 7 12 If overall threshold checking is turned off the tqzip program displays the following prompt gt Threshold checking is OFF gt Do you want to turn it ON y lt n gt gt y gt gt Threshold checking is now ON DHAKA KKK KKK KK KK KKK KK KKK KK KKK KKK KKKKKKKKKKK KKK KKK KK x TQALM must be executing on your system gt when threshold checking is ON DAK K KKK KKK KK KK KKK KK KKK KKK KKK KKK KKKKKKKKKK KKK KKK KK Overall threshold checking must be turned on before any of the individual threshold definitions are checked for threshold violations To turn on overall threshold checking you can either answer y to this prompt or use the Turn Threshold Checking On Off selection on the Alarms and Threshold Menu to perform this task see 7 3 When using threshold checking you must also ensure that the tqalm program is executing on your system See 7 17 for information on the tqalm program and alarm handling TQ 01011 18 7
181. ame specifies the name for the table File_Name file name Specifies the name of the file in which to store the table and is a required item File_Name is the keyword for the item file name is the file name in which to store the table TQ 01011 18 9 49 Customizing Data Collection Probes time type Can be one of the following optional See 9 11 2 5 for a description of these time types Table_Time_Granulized Table_Time_Sampled Table_Time_ Event Retention ndays Specifies the retention period for the primary retention A primary retention specification is automatically created when the table is created If this optional keyword is specified the primary retention period is set to ndays Otherwise it is set to 1 day Example For the example 2200 is the table class accountdisk is the table name and usertabs is the file name The table definition input is as follows TABLE Table_Name 2200 accountdisk File_Name usertabs The file TQDATA dbusertabs will be created on the system This file should be added to the set of files that are secured on the system Field Definition Guidelines Defining TeamQuest Baseline tables may appear complex when looking at the various field definition options In reality there are some basic guidelines that you can follow to make the process easier and that should address most requirements e Group Key Fields Each table must have one or more group keys for record sequencing and indexing These f
182. an be a string of up to 52 characters which defines the Keyname The keyvalue specifications must be enclosed in quotation marks 1 include xxx exclude xxx C 30 When this keyword is used within a list of keywords you will be prompted to supply information to complete the field at the time the report is opened For example if the appears as the keyvalue for the Resource TeamQuest View displays a list of all legal resources specified for statistic reported You will not be prompted and the field will be automatically filled in if there is only one choice This keyword is like the keyword except it will be always saved in the report definition regardless of the save preferences set in the Miscellaneous settings of User Preferences This keyword corresponds to the Always Ask button in the prompt dialog boxes This keyword is like the keyword except you will be prompted even if there is only one choice This is used to allow defining other selection settings such as include exclude and previous group settings even when there is only one choice in the database you are currently working in These keywords are used to define an include mask pattern It is used to include choices from a list based on a mask statement or an exclude mask pattern The patterns can be made up of any character with the following mask characters available Matches zero of more choices Matches exactly
183. ance Database The TeamQuest Baseline probes obtain performance data from the server system and store the data in the performance database The performance database consists of data values for the parameters gathered from the various collection sources Data is collected and stored in the database according to defined data aggregation sets Aggregation sets parameters and data values are concepts essential to the understanding of the performance database and the use of TeamQuest View to retrieve and display data from the database Table data is also maintained in the performance database at the interval where it was originally collected Like aggregation set data table data can be retrieved and displayed by using the TeamQuest View Aggregation Sets The TeamQuest Baseline lets you collect and store your performance data at different sampling intervals or levels of granularity Depending on your needs you may be interested in data sampled at a small granularity such as 1 minute intervals or at a larger granularity such as 1 hour intervals TeamQuest Baseline and TeamQuest Online use aggregation sets to define the granularity of the data in the database as well as the length of time the data is retained in the database TeamQuest Online allows only one aggregation set in the database Data records are saved into a 1 minute aggregation set This means that data is collected at 1 minute intervals and is retained in the database for a period of
184. and Total SUPs used to process the job Sum of the IP SUPs ER CC SUPs operating system SUPs and I O SUPs Total IP SUPs Total ER and control card SUPs Total I O SUPs Total Exec Services I O SUPs Total Exec Services IP SUPs Total control statement I O SUPs Total control statement IP SUPs Total core block SUPs Sum of SUPs and main storage usage Total number of I O requests made by this run Total number of I O requests to disk that have been made by this run Total number of words that have been transferred during disk I O operations by this run Total number of I O requests to tape that have been made by this run Total number of words that have been transferred during tape I O operations by this run Total voluntary wait time Sum of user response time and TWAIT time Total amount of time the run spent waiting for a resource Total time the job was active Total number of cards the Exec read during the run Total number of runs that were opened Maximum number of real pages concurrently resident for this run The product of the total number of resident pages and the time in seconds that the pages are resident TQ 01011 18 OS 2200 Parameters Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs ES IO SUPs ES IP SUPs CS IO SUPs CS IP SUPs IO SUPs Core Block SUPs IO Requests Disk I O Requests Disk I O Words Tape I O Requests Tape I O Words Voluntary Wa
185. and alarm handling It also outlines steps you can take to ensure that you are collecting the data you need for your site requirements TQ 01011 18 About This Guide Part 2 Customizing Your Site Section 4 TeamQuest Performance Software Administration This section provides information to establish and terminate an administration session TeamQuest Performance Software Administration allows you to perform maintenance tasks for workloads aggregations and derived statistics Section 5 Customizing Workloads This section provides information on the workloads used by TeamQuest Baseline It describes workload characteristics and the definition language for defining workloads It also tells you how to use the tqzip program to unload and load workload definitions to an output file Section 6 Customizing Aggregation Sets This section contains advanced concepts about the aggregation sets in the performance database and tells how you can customize the aggregation sets to meet the particular needs of your site It provides information for using the tqasa and tqagg programs to perform tasks for maintaining aggregation sets Section 7 Customizing Alarms and Thresholds This section provides information on the alarms and thresholds capability of TeamQuest Baseline It tells you how to perform the various tasks in creating and maintaining alarms and thresholds for database parameters and derived statistics Section 8 Customizing Derived S
186. and the unit If the keepcount parameter is not specified 180 groups is assumed The maximum retention period that can be selected is dependent upon the granularity of the aggregation set being defined based on a maximum of 31 680 data points that can be stored in 1 day for any aggregation set Some common granularities and their maximum retention periods that can be selected are Granularity Maximum Retention Period 15 seconds 5 days 30 seconds 11 days 1 minute 22 days 5 minutes 110 days 10 minutes 220 days 30 minutes 660 days 1 hour 1320 days specifies the aggregation set that feeds the aggregation set being created The fed by parameter is the name of an aggregation set that has already been defined for the database You must use the fed by parameter if you want this aggregation set to receive data from another aggregation set in the database If you do not specify the fed by parameter the aggregation set receives data only from the TeamQuest Baseline probes Note that the TeamQuest Baseline probes support only day group types specifies the days of the week to exclude for the aggregation sets that have week and month group types specified The exclude parameter has the format of xxxxxxx where x can be either 1 or 0 A value of 1 signifies a day to exclude A value of 0 signifies a day to include This exclusion mask is a series of seven 1 and 0 digits with each digit representing a day of the week The first digit represen
187. ank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec TQ 01011 18 Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load Demand ZOOMs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load Batch Abs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load Batch ZOOMs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second A 19 OS 2200 Parameters A 20 The following statistics are available from Exec level SB4R3 through level SB5R4 Parameter Hierarchy Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads S
188. ansactions that may be held up for some particular reason This information is reported in a tabular column format and is displayed by TeamQuest View See A 9 2 for a list of the information stored in the Transactions table by tqsam TQ 01011 18 9 5 9 5 1 Customizing Data Collection Probes MIPS Metering Data Collection TeamQuest Baseline contains a MIPS Metering data collection runstream called tqmips The tqmips runstream uses a run identifier of tgmips and executes the tqmips program The tqmips program reads from the dedicated Capacity on Demand COD audit trail or the system log file and collects MIPS Metering data for a COD partition and stores it in the performance database The data collected by tqmips is also available for real time system monitoring You should allow the tqmips runstream to execute continuously as a background run on your system This ensures that your performance databases contain a complete set of data for each day and that the real time monitor can access the most current data tqmips Processor Call Statement The tqmips runstream contains the tqmips processor call statement which controls the collection of MIPS metering data collection There are two trails that contain the same MIPS information By default the tgmips probe reads and processes from the dedicated COD audit trail The COD audit trail contains only COD MIPS related data MIPS data is also contained in the system log but it is much le
189. ant to deregister at the prompt The tqzip program then provides a message indicating the table that has been deregistered The 2200 Archive Runs table is deregistered The table name is removed from the list of tables in the database TQ 01011 18 Customizing Tables 11 10 Unloading Table Specifications The tqzip program allows you to export data collection and retention specifications for the tables currently contained in the performance database into a file You can use this file to view the data collection and retention specifications and update them by using a text editor You can then reload the table specifications by using the l and c options on the tqzip command line see 11 11 To unload table specifications with tqzip use the following command line format tqzip d databasename u c tablespec file where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting of the tq ini element is used see 3 1 u specifies that an unload operation is to be performed by tqzip The type of unload operation is table collection retention as indicated by the c option C unloads the data collection and retention specifications from the database to the file specified by the tablespec file parameter Example The following shows an example of unloading table specifications to the file
190. ap variables are to be sent See J 4 2 for a list of the trap variables By default all trap variables are sent with each alarm trap However only the minimum set of trap variables is displayed at the management console To have the additional trap variables displayed you must update the configuration information for your management console product TQ 01011 18 2200 SNMP Trap Alarms J 5 1 HP OpenView Configuration To make your TeamQuest Performance Software alarm display a message on an HP OpenView Event Browser window you must first define a threshold using the tqzip program and then configure the HP OpenView trap daemon to accept and process the trap Follow these steps 1 Define a threshold using the tqzip program and add the SNMPTrap command to one of the User Text fields in the threshold definition Since you can select fields in the HP OpenView trapd configuration you do not need to specify which fields to send on the SNMPTrap command On the SNMP MGMT statement in the CMS 1100 configuration you must specify the host that has the HP OpenView trap daemon trapd or ovtrapd running on it Also the trap daemon must be listening on port 162 which is the destination port for all CMS 1100 SNMP traps 2 Configure the HP OpenView trap daemon trapd or ovtrapd to accept and process your alarm trap To do this merge the appropriate TeamQuest supplied file into the trapd conf file as documented in the trapd man pages These en
191. ard deviations than with compact aggregation sets This is because information is saved with each data point allowing the standard deviation to be calculated at report time Also there is increased accuracy for very large or very small data point values since double precision floating point is used to store the values in an expanded aggregation set Note Before level 6R2 all aggregation sets created using TeamQuest View or the tqasa program were in expanded format You can convert existing expanded aggregation sets into compact aggregation sets by using the tgasa program not available with TeamQuest Online This conversion will result in a 40 to 60 reduction in the disk space required by the aggregation set See 6 4 6 for information on how to convert expanded aggregation sets into compact aggregation sets Maintaining Aggregation Sets with TeamQuest Performance Software Administration To perform any administration tasks using TeamQuest Performance Software Administration you must have a session established to the host software see 4 2 You access the dialog boxes for performing administration tasks by selecting the Administration menu Note When aggregation set maintenance tasks are being performed the database is locked Only the user performing the tasks can access the database for creating new aggregation sets Other users can view aggregation set information in read only mode when the database is locked TeamQuest Performance
192. are retrieved from the alarm history table Alarm messages can consist of up to four lines of text The first line is generated by TeamQuest Baseline The next three lines of text are the user alarm message text lines that you specified when creating the threshold definition for the parameter see 7 4 The alarm message for a parameter has the following format TQ severity lt Xxxxxx xx gt system alarm id quall qual2 qual3 user alarm message line 1 user alarm message line 2 user alarm message line 3 where severity is NORMAL WARNING or CRITICAL WARNING and CRITICAL indicate the raising of an alarm NORMAL indicates the clearing of an alarm See 2 6 for more information on alarm severities lt XXXXXX XX gt is the value of the parameter when the alarm occurred system is the system name on which the alarm occurred alarm id is the alarm identifier specified in the threshold definition of the parameter quall is the resource and or application group name for the parameter qual2 is the workload set for the parameter TQ 01011 18 Customizing Alarms and Thresholds qual3 is the workload for the parameter user alarm message is the user defined text as specified in the threshold definition line 1 2 3 for the parameter The alarm message can contain up to three lines of user alarm message text These lines are optional and are not displayed with the alarm when the severity is NORMAL Example The following alarm messa
193. art time can be any time of the day but it must be an even multiple of the minimum collection period for the probe 11 6 6 Exiting the Retention Specification Menu When you exit the Retention Specification Menu you can apply any retention specification changes you made by selecting Exit Retention Specification Menu Enter 99 at the prompt If you do not want to apply the retention specification changes you can select Omit Retention Specification Changes from the menu Enter 98 at the prompt Retention Specification Menu 2200 Runs List Retention Specifications Modify Primary Retention Specification Add a Retention Specification Delete a Retention Specification Set Retention Start Time 98 Omit Retention Specification Changes 99 Save Changes and Exit Retention Specification Menu SMS NPN ENE ME MOME NONSENSE N arwWNrMrHr gt gt Enter your selection lt 1 gt gt 99 When you select 98 or 99 at the prompt the tqzip program returns you to the Tables Menu see 11 8 TQ 01011 18 11 15 Customizing Tables 11 7 Deleting Records from Tables Normally records are deleted from tables by the TeamQuest Baseline aggregator based on the retention specifications for each table However there may be times when you want to remove data records manually by using the tqzip program To begin record deletion for a table select Record Deletion from the Tables Menu see 11 3 by entering 4 at the prompt
194. assumed to be Day By definition a day ends at 24 00 00 and a week ends on Saturday at 24 00 00 RelativeDate Is a relative integer which represents a CalendarType relative to the current date If CalendarType is Day 0 represents today 1 represents yesterday 2 represents the day before yesterday and so on If the CalendarType is Week 0 represents the current week 1 represents last week and so on View Spread Sbar Pie Cluster Line Area Point Scatter Trend The type of graph or text report that TeamQuest View displays for the report Only one type of report can be specified in this field The default value for this field is Line The following types of reports can be specified Spread Displays data as textual reports in spreadsheet format Sbar Displays data in stacked bar graph format Pie Displays data in pie chart format Cluster Displays data in cluster bar graph format Line Displays data in line graph format C 6 TQ 01011 18 Report Definition File Area Displays data in area graph format Point Displays data as points only Scatter Displays data in scatter plot format The data series and line type settings for the x axis and y axis are defined in the Scatter key Trend Displays data in a bar graph format with a trend line The series and line types are defined in the Trend key Format XAxis LegendEntries Page TQ 01011 18 The specifications for the grouping of data along the axis This
195. at reduce the level of detail required to understand the operation of the computer system and its relationship to the various business operations it supports This process is essential to understanding how well your computer system is meeting the requirement of providing acceptable levels of service to the users TQ 01011 18 TeamQuest Baseline and TeamQuest Online Concepts Workload characterization allows you to identify the workloads and track their use of system resources Each workload is uniquely identified and defines a criteria for distributing the system resources used by each activity For example you can define your workload to track how much of the system resources are used by payroll order entry or a software development project The system activity monitor TPM IOTRACE MAPPER log and system log probes use the workload definitions to distribute the statistics for the work performed by each activity to the appropriate workload TeamQuest Baseline supports the definition of multiple workload sets each containing unique workload definitions A workload set is defined as a set of workloads that represents the system resources used by all activities on the system TeamQuest Baseline uses each workload set for distributing the work performed by each activity This allows you to track the activities on the system using a different view of the system Creating multiple workload sets is useful for testing new workload definitions becaus
196. at contains the aggregation set you want to convert The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 Example The following example shows how to convert an aggregation set from expanded to compact format gt tqasa c A gt TQASA 7R4G 011312 11 04 45 Fri Jan 13 16 22 10 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt DH KK KKKKKKK KKK KKK KKK KKK KKK KKK KK KKK KK KK KKK KKK KKK KKK KKK KKK B gt Aggregation set compaction has been selected gt for database lt PRODUCTION gt gt gt It is recommended that you backup your database gt before continuing with this operation gt gt All processes currently using the selected gt database must be stopped before continuing DH KKKKKKKKK KKK KKK KK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK gt Do you want to continue y lt n gt C gt y gt Enter the aggregation set name D gt l minute gt gt Compaction is complete for aggregation set 1 minute E TQ 01011 18 6 4 7 Customizing Aggregation Sets The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A Call the tqasa program with c option to convert your aggregation sets B The tqasa program displays messages for your
197. at the tqalm program is executing on your system See 7 17 for information on the tqalm program and alarm handling 7 10 TQ 01011 18 Customizing Alarms and Thresholds 7 5 Reusing Threshold Definitions When you are defining a threshold for a parameter that has a resource workload or application group associated with it the tqzip program allows you to use the same threshold definition for other resources workloads or application groups During the threshold definition sequence the tqzip program displays the appropriate prompts depending on the parameter The following example shows the prompt for a resource parameter These prompts are displayed before you are asked to create or update another threshold see Q in 7 4 gt Define thresholds for another resource lt y gt n A gt y gt gt Resource Name gt 1 DSKAO gt 2 DS gt 3 DS gt 4 DS gt 5 DS gt 6 DS gt 7 DS gt 8 DS gt 9 DS 0 NODOVUD1NDWDOYS gt We AANA IFWMPMHe H SOS NE NSE SOS ONS SG gt Enter number name lt blank gt for more or E to exit B gt CLO1 10 Activity Disk Summary Percent Busy DSKAI1 C gt 1 Threshold checking ON gt 2 Threshold 1 00050 00 Threshold 2 00075 00 gt 3 Number of zones e253 Normal Zone sol gt 4 Secs before alarm 0060 Secs delay 0060 gt 5 Alarm identifier Percent Busy gt 6 User text When more than 50 busy call John at ext 2332 gt 7 User
198. ata series If both EnableLow and EnableUp are set to Off the threshold box does not appear on the report legend The default for this field is Off Upper Specifies the upper threshold limit When parameter values are greater than or equal to this value the threshold box for the data series appears in the color or pattern specified in the Exception field of Threshold Color or Threshold Pattern as defined in the general section C 34 TQ 01011 18 Report Definition File Parameter On Off Multiplier Identifies if this parameter is currently displayed and what multiplication factor is applied to all values of this parameter The acceptable values are On or Off On indicates that the parameter is displayed Off indicates that the parameter is hidden Multiplier Specifies the multiplication factor applied to all values of the parameter Lower Threshold Process executable name The alarm program that is specified for the lower threshold limits for the parameter values of this parameter When parameter values exceed the specified limits the alarm program is invoked according to the values specified in the Sequence statement of the general section Upper Threshold Process executable name The alarm program that is specified for the upper threshold limits for the parameter values of this parameter When parameter values exceed the specified limits the alarm program is invoked according to the values specified in the Seque
199. ation Defines the number of points for the typeface of the labels This value determines the height of the characters in which the labels are printed The value must be a positive integer Report Definition File Color Specifies the color in which the vertical axis and the labels are printed The default value is black The default values for Value Axis are defined in the Axis Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Value Line Enable Thick MajorTick MinorTick MajorGrid MinorGrid The specifications for the line attributes of the vertical axis of the report Enable Indicates whether the vertical axis line appears on the report The acceptable values are ON or OFF If Enable is set to ON the vertical axis line appears on the report If Enable is set to OFF the vertical axis does not appear on the report The default for this field is ON Thick Defines the weight in pixels of the vertical axis line The value for this specification must be a positive integer The acceptable values are 1 2 and 3 The default value is 1 MajorTick Defines the type of marks used for the major ticks on the vertical axis line The acceptable values are Invisible No ticks marks appear Inside Tick marks appear on the inside of the axis line Outside Tick marks appear on the outside of the axis line Crossing Tick marks cross the axis line The default value for the M
200. ation also identifies the name of the performance database you are using By creating a session definition and saving this information under a session name you can reuse it for subsequent queries of that data When you create a new session you will specify how to connect to the host software where your performance database exists You can connect to the host software using a socket or named pipe To create a new session 1 2 TQ 01011 18 Select File gt New Session The New Session dialog box is displayed This dialog box is used to select the type of communications connection to access the performance database New Session Host Name or IP Address e101 Connection Socket SSL Named Pipe From the Connection option buttons select the type of connection you want to use for the session you are defining When you select the appropriate option button for your connection the fields on the right side of the dialog box will change to allow you to provide the information required to make the connection Socket For more information see 4 4 1 SSL For more information see 4 4 2 Named Pipe For more information see 4 4 3 Enter the information required to make the communications connection Click the OK button The New Session dialog box closes and TeamQuest Performance Software Administration initializes communication with the host software Messages indicating the progress of the initializa
201. ation Set Definitions 6 4 6 Converting Expanded Aggregation Sets _ 6 4 7 Performing Data Aggregation oo eeeeeeeeneees Customizing Alarms and Thresholds 7 1 INWOCUCIION sarsaran ie ania Heine 7 2 Concepts apoien 4n i dati Ae ba ee 7 3 Opening the Alarms and Thresholds Menu _ 65 7 4 Creating Threshold Definitions eseeeeeeeseeeeeeeeeen 7 5 Reusing Threshold Definitions seenen 7 6 Updating Threshold Definitions n se 7 7 Viewing Thresholds oo eeeeeeeeceeeeeeeeceeeeeeeeeeeeeeeteeeeeenees 7 8 Creating Derived Statistic Threshold Definitions _ 7 9 Updating Derived Statistic Threshold Definitions 7 10 Viewing Derived Statistic Threshold Definitions __ 7 11 Deleting Thresholds o eeeceeceseeeeeeeeneeeeeeeeeneeeeeeeeaaes 7 12 Notifying Probes of Threshold Changes _ 0 08 7 13 Unloading Threshold Definitions oes 7 14 Loading Threshold Definitions oo eeeeeeeeeeeeeee 7 15 Turning Threshold Checking On and Off n se 7 15 1 Using the Alarms and Thresholds Menu _ 7 15 2 Using the tqzip Processor Call sese PPPRP NNAAN ct DDD MD MD Phi ono OO TORE PMD A OO 6 26 6 ro N 7 1 7 1 7 5 7 6 7 11 7 12 7 13 7 14 7 16 7 18 7 19 7 21 7 21 7 24 7 25 7 25 7 26 xiii Contents xiv Section 8 Section 9 7 16 7 17 7 18 7 19 7 20 Generating Alarm Lists oo eceeeeeeceeeseeeee
202. atistic Name Reqs Sec Words Req Avg RET Response Time Category Group Category Resource Subcategory Statistic Name Reqs Sec Words Req Avg RET Response Time IO Activity XPC XPCO XPC1 Write Miss XPC write miss requests per second XPC write miss words per request XPC write miss average request existence time XPC write miss average response time IO Activity XPC XPCO XPC1 Destage XPC destage requests per second XPC destage words per request XPC destage average request existence time XPC destage average response time IO Activity XPC XPCO XPC1 Bypass XPC bypass requests per second XPC bypass words per request XPC bypass average request existence time XPC bypass average response time IO Activity XPC XPCO XPC1 Misc Exec XPC miscellaneous Exec requests per second XPC miscellaneous Exec per request XPC miscellaneous Exec request existence time XPC miscellaneous Exec response time TQ 01011 18 OS 2200 Parameters Category Group Category Resource Subcategory Statistic Name Programs Sec Avg Queue Length Program Rsp Time Lock Reqs Sec Lock Process Time A 2 7 Queue Statistics IO Activity XIIP XITPO XIIP1 Summary XPC programs per second on this XIIP XIIP initiation queue length XPC response time for programs from this XIIP XPC L lock function requests per second from this XIIP Average XPC L lock command processing time in milliseconds from this XII
203. be considered the first day of the week TeamQuest Baseline normally assumes that Sunday is the first day of the week unless you specify another day as the first day of the week You use TeamQuest Performance Software Administration to create the aggregation sets for data consolidation To consolidate data from one aggregation set to another you must use the tqagg program see 6 4 7 Figure 6 2 illustrates the concept of consolidating data from one aggregation set to another In this example the performance database has two aggregation sets defined each containing information collected 24 hours each day The first aggregation set is the same as shown in Figure 6 1 It contains data records saved for every parameter at 30 minute intervals a total of 48 data records for each parameter of each day The second aggregation set saves the data records at 1 hour intervals weekly for 52 weeks In this aggregation set TeamQuest Baseline saves a total of 24 data records for each parameter summarized by week At the end of each week TeamQuest Baseline automatically rolls the data from the first aggregation set into the second aggregation set consolidating seven daily groups each with 48 time groupings into one weekly group with 24 time groupings TQ 01011 18 6 5 Customizing Aggregation Sets Aggregation Set 1 gt Aggregation Set 2 Time Grouping Time Grouping Date Grouping
204. bed in this appendix e The text that appears in normal type must be entered exactly as shown However you can use either uppercase or lowercase characters e Italics indicate fields or values that you must supply e Multiple choices are separated by commas You must type the comma between fields e Parentheses indicate optional choices Do not enter the parentheses e Some terms are enclosed in braces and separated by a vertical bar You must choose one of the terms enclosed within the braces Do not type the braces e Section headers must be enclosed within brackets within the report definition e The comma is the continuation character If the syntax does fit on a single line complete the field end the line with a comma and continue on the next line e For any field leading and trailing spaces are ignored If a field contains spaces it must be enclosed in quotation marks e Fields enclosed in quotation marks cannot contain commas C 3 File Format A report definition file is divided into sections with each section defining a specific aspect of the report Each section contains a section header followed by the definition items that provide the specifications for the report The sections for a report definition file are e General this section defines the attributes for the layout and object attributes of the report TeamQuest View uses the specifications in this section for the initial presen
205. by workload deadline batch population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per deadline batch run demand rpt Demand Workload Performance Using statistics derived from the tqsys probe this report shows by workload demand population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per demand run TQ 01011 18 Predefined Report Descriptions File Name Description File Name Description File Name Description TQ 01011 18 tip rpt TIP Workload Performance Using statistics derived from the tqsys probe this report shows by workload TIP population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per TIP transaction hvtip rpt HVTIP Workload Performance Using statistics derived from the tqtpm and tqrtpm probes this report shows by workload HVTIP population throughput response time CMS queue time processing time lock wait time termination time IP SUPs per second ER CC SUPs sec IO SUPs sec FCSS reads per second and FCSS writes per second mapper rpt MAPPER Workload Performance Using statistics derived from the tqmpr probe this report shows by workload MAPPER population
206. c Total Exec MIN Total Exec MAX Total User Total User MIN Total User MAX Misc NonSwitchable IO Processing Processor Control Memory Management TIP Integ Rec Aud Cntrl Symbionts Misc Switchable Real Time ILES CMS 1100 UDS Processor Java Total Collection Base Total available time for all IP performance monitors for this sample Percent all IP performance monitors were available for this sample Processor Java Total Utilization Percent of time idle Minimum percent of time idle Maximum percent of time idle Percent of time in Exec processing Minimum percent of time in Exec processing Maximum percent of time in Exec processing Percent of time in user processing Minimum percent of time in user processing Maximum percent of time in user processing Percentage of miscellaneous nonswitchable Exec processing Percentage of Exec I O processing Percentage of Exec processor control processing Percentage of Exec memory management Percentage of Exec TIP integrated recovery and audit control processing Percentage of Exec symbionts processing Percentage of miscellaneous switchable Exec processing Percentage of user real time processing Percentage of independently linked Exec subsystems Percentage of CMS 1100 processing Percentage of UDS processing TQ 01011 18 OS 2200 Parameters TIP TIP MIN TIP MAX Critical Deadline Demand Batch Category Group Category Subcategory Statist
207. c name The coarser point is a weighted sum of the finer points and the value of the statistic name is used as the weight for computing the weighted sum Given a sequence of values and weights V1 Vo V3 Vn and Wy Wo W3 Wy the weighted sum is computed as vy wy Vo Wo V3 Wg Vy Wy wy Wo Wg Wp The value of the statistic name is used as the weight for the corresponding field NONE For reserved fields such as DATE TIME NONE and RESOURCE you must specify a weight of NONE This is used as a place holder to record something in the weights line that corresponds to the reserved word fields TQ 01011 18 Customizing Data Collection Probes PRBDATA File Input Format The data to be stored into the performance database is found in the file with an internal use name of PRBDATA This file must correspond in format to that described by the first line of the TQARG file This file is created by a user written process This process can be an addstream that executes the various programs needed to create the output in the needed format The following is an example of a PRBDATA file This PRBDATA file could be used with the TQARG file described previously 01 14 11 16 31 00 36855 11905 21270 36 areal 01 14 11 16 31 00 249991 206490 18511 92 area2 01 14 11 16 31 00 125247 19166 93561 17 area3 01 14 11 16 31 00 134839 104380 16979 86 area4 The lines of the PRBDATA file must contain the actual data samples that the tq
208. category or statistic of the item that is selected Expand button allows you to expand the hierarchy list of a selected key item to the next level Apply button allows you to send an alias selection back to the Derived Statistic Definition dialog box while staying in the Define Alias dialog box to define more aliases View button allows you to see the key values that are truncated in the list box see 8 3 5 TQ 01011 18 8 11 Customizing Derived Statistics All of the aliases defined using the Define Alias dialog box refer to unqualified statistics The statistic itself may have relationships that must be defined to retrieve data for the statistic from the database Explicitly specifying values for these relationships is called qualifying the statistic A fully qualified statistic is one whose relationships are all defined such that a value could be retrieved from the database using the given information For TeamQuest Baseline all statistics need a system qualifier In addition some statistics need to be qualified by resource or by workload set and workload or by application group The qualification does not need to be explicitly specified because it can be obtained at report time For all of the aliases defined using this dialog box required qualifiers are specified using a substitution reference SYSTEM RES WLS WL or AG The substitution reference tells the server that these values are filled in at report time The deri
209. cation for the lower coordinate of the bottom right hand corner of the annotation box Right Specifies the relative location for the right coordinate of the bottom right hand corner of the annotation box Vertical Defines the vertical alignment of the text within the annotation box The acceptable values are Top Aligns the text at the top of the annotation box Center Centers the text vertically in the annotation box Bottom Aligns the text at the bottom of the annotation box Horizontal Defines the justification of the text within the annotation box The acceptable values are Left Justifies the text to the left in the annotation box Center Centers the text horizontally in the annotation box Right Justifies the text to the right in the annotation box Font Specifies the typeface in which the annotation text is printed on the report TQ 01011 18 C 25 Report Definition File Style Specifies the kind of emphasis for the typeface of the annotation text Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the annotation text be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Specifies the height of the characters measured in points of the typeface in which the annotation te
210. ce data see 9 11 1 The tqusrprb program stores data into the specified aggregation sets of the performance database The data stored in an aggregation set consists of a real number value its weight and its sum of squares for each statistic Examples of such statistics are CPU usage disk space usage network utilizations and so on The data from aggregation sets can be charted and analyzed using TeamQuest View in the same way that data from tqbsp is charted and analyzed e Use the tqtblprb program when collecting table data see 9 11 2 The tqtblprb program stores data into tables in the performance database rather than into aggregation sets You will want to use table storage for such data as process data where there are varying numbers of data items at each interval with nonnumeric data command name user name group name and so on or alarm data where the data is event based rather than time based You can analyze the data using TeamQuest View This data can only be viewed as tables it cannot be viewed as charts Note When you are experimenting with user probes do not work on your primary database as you may create several experimental statistics that can cause your aggregation sets to expand unnecessarily You should create an additional database using the tqgdbu program see 10 4 and experiment with your new user probes on a test database first Only when you are sure of the user probe should you use it to store data in the pr
211. collection description 9 33 file input to tqsys 9 35 parameters A 46 reports B 15 specifying workload definitions 5 1 starting 9 383 System Table Open Distributed Transaction Processing A 61 T table data description 2 4 host based reporting E 1 storing 9 57 tqsam probe 9 8 Index 6 tables adding collection period 11 7 retention specifications 11 13 administering collection periods 11 4 applying collection period changes 11 9 retention specification changes 11 15 Client A 64 collection periods 11 1 deleting collection periods 11 8 records 11 2 11 16 retention specifications 11 14 deregistering 11 20 displaying 11 4 listing collection periods 11 6 retention specifications 11 11 main menu 11 3 maintaining 11 1 11 3 modifying primary retention specification 11 12 registering 11 2 11 18 reorganizing table files 11 26 retention periods 11 2 Runs A 58 Server A 62 Service A 64 setting retention start time 11 15 specifications file format 11 21 System A 61 tqzip administration 11 3 Transactions A 59 turning on and off data collection 11 9 Tables Menu of tqzip 11 3 TeamQuest Baseline aggregation set description 3 3 alarm handling 3 3 capabilities 1 5 considerations for using with SPO 7 32 data collection 1 6 data collection control 9 1 data display 1 6 database 1 10 2 1 default threshold definitions changing 3 5 description 3 4 using 3 4 description 1 4 host software description 1 8 TQ 0
212. cons method field name Non_Seq_Cons_Method cons methodl field name where FIELD Defines a field input line and must be first on the line This item is required Field_Name field name Specifies the field name and is a required item Field_Name is the keyword for the item field name is the name for the field Group_Key Indicates that the field is part of the group key for the table Normally the sample time sample interval and system name make up the group key If Group_Key is not present the field is not part of the group key TQ 01011 18 9 51 Customizing Data Collection Probes Primary_Key Indicates that the field is part of the primary key for the table Normally the fields that are group keys and any additional fields required to uniquely identify arecord are primary keys If Primary_Key is not present the field is not part of the primary key Identifier Indicates that the field is an identifier for the table An identifier describes the object being measured rather than containing a value If Identifier is not present the field is not an identifier field field type Can be one of the following optional See 9 11 2 5 for a description of these field types Field_Type_None Default Field_Type_Timestamp Field_Type_Desired_Interval Field_Type_System data type Can be one of the following optional See 9 11 2 5 for a description of these data types U_Long_Int_Type default Long_Int_Type
213. contingency it restarts automatically If tqodtp terminates and does not restart try to resolve the problem as soon as possible and restart tqodtp Data is collected only when tqodtp is running TQ 01011 18 9 15 Customizing Data Collection Probes 9 6 4 9 6 4 1 9 6 4 2 9 6 4 3 9 6 4 4 9 16 tqodtp Table Data The tqodtp probe collects and stores data into several tables The tqodtp probe monitors ODTP information using an interface supplied with the OLTP TM2200 product The tqodtp probe monitors the information at each sample interval The ODTP configuration parameter STATS_LEVEL must be set to one or greater for the data to be available for collection by the tqodtp probe For additional information on configuring STATS_LEVEL see the Distributed Processing Middleware Open Distributed Transaction Processing Administration Guide Volume 1 Installation Configuration and Ongoing Administration System Table Information The System table contains information specific to the ODTP environment This information is reported in a tabular column format and is displayed by TeamQuest View See A 10 1 for a list of the information stored in the System table by tqodtp Server Table Information The Server table contains information specific to the ODTP environment This information is reported in a tabular column format and is displayed by TeamQuest View See A 10 2 for a list of the information stored in the Server table by tq
214. cription 9 29 fileinput totqmpr 9 31 parameters A 45 reports B 14 specifying workload definitions 5 1 starting 9 29 memory reports B 7 statistics A 18 MIPS metering data collection description 9 9 starting 9 9 modifying aggregation set 6 14 derived statistic 8 16 primary retention specification for tables 11 12 workload definition 5 16 workload set definition 5 11 N named pipe connection defining 4 6 Index 4 O Open Distributed Transaction Processing Client table information A 64 description of data collection 9 14 Service table information A 64 starting data collection 9 14 System table information A 61 P paging statistics A 22 parameter data value 2 3 description 2 2 hierarchy of structure 2 2 key names 2 2 listing 10 4 parameters Capacity On Demand systems A 43 IOTRACE A 44 MAPPER log A 45 MAPPER log probe A 53 Open Distributed Transaction Processing A 61 SIP A 2 SIP probe A 51 system activity monitor A 58 system log A 46 system log probe A 54 TPM log A 41 parameter group section C 29 password encryption I 10 performance database backup and recovery 10 8 performance monitor statistics A 9 performing data aggregation 6 27 derived statistic maintenance 8 5 probe description 1 8 tqbsp 1 8 9 2 tqiot 1 9 9 25 tqmips 9 9 tqmpr_ 1 9 9 29 tqodtp 1 10 9 14 tqrtpm 1 9 9 17 tqsam_ 1 9 9 5 tqsys 1 10 9 33 tqtblprb 1 10 9 38 TQ 01011 18
215. cription of aggregation sets p specifies a MAPPER name The mapper name parameter is the name that is associated with the MAPPER environment that generated the accounting log file It is used to determine where to store the data in the performance database The mapper name is case sensitive Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent lines of the processor call The parameters specified on the processor call must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keyword The tqmpr program processes each MAPPER accounting log file storing data points in the performance database as specified by the aggregation set parameter a TQ 01011 18 Customizing Data Collection Probes 9 9 3 MAPPER Log File Input With the tqmpr program you can load data from a MAPPER log file into a performance database by specifying the specific MAPPER log file This enables you to load MAPPER log data from another system or to update the performance database with MAPPER log data that was collected at a time when tqmpr was not running on your system To load data directly from specific MAPPER log files you must specify two additional parameters on th
216. ct this value when the report is opened Statisticltem Is a keyword chosen from the Statistic keyword list A can be specified instead of an integer value to allow you to select this value when the report is opened mu Is for backward compatibility with previous release levels Leave this setting as a blank string or it may be set to the reserved keyword to cause the interface to prompt you to modify Top N settings LastPoint If set to On only the last point is used to determine the Top N order If set to Off or not set the entire report calendar range is used to determine the Top N order Include the reserved keyword to cause the interface to prompt you to modify Top N settings TQ 01011 18 C31 Report Definition File Bottom Positive integer StatisticItem LastPoint The specifications for selecting the least active resources for a given statistic Positive integer Specifies the number of the least active resources that are displayed A can be specified instead of an integer value to allow you to select this value when the report is opened Statisticltem Is a keyword chosen from the Statistic keyword list A can be specified instead of an integer value to allow you to select this value when the report is opened wo Is for backward compatibility with previous release levels Leave this setting as a blank string or it may be set to the reserved keyword to cause the interface
217. ction for XPA systems TIP tape I Os per second TIP disk I Os per second TIP I O SUPs per second TIP ER CC SUPs per second TIP IP SUPs per second Seconds per TIP transaction TIP transactions per second Average number of active transactions TQ 01011 18 OS 2200 Parameters Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec req Throughput reqs sec Population open runs Category Group Category Subcategory Workload Set Workload Statistic Name Kmem run XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec req Throughput reqs sec Population open runs TQ 01011 18 System Log Overall Demand WLS1 WLS2 K words of memory per demand run for XPA systems Demand tape I Os per second Demand disk I Os per second Demand I O SUPs per second Demand ER CC SUPs per second Demand IP SUPs per second Seconds per demand request Demand requests per second Average demand runs open System Log Workload Demand WLS1 WLS2 all K words of memory per demand run for XPA systems Demand tape I Os per second Demand disk I Os per second Demand I O SUPs per second Demand ER CC SUPs per second Demand IP SUPs per second Seconds per demand request Demand requests per second Average demand runs open A 55 OS 2200 Parameters
218. ctions under runs Number of input operations from functions under runs Number of user terminal interactions that occurred Number of seconds per interaction Number of user terminal interactions during runs that occurred A 45 OS 2200 Parameters Non Run Interactions Number of manual user interactions that occurred Response Sec Run Int Number of seconds per run interaction Response S Non RunInt Number of seconds per manual interaction A 7 System Log Parameters A 46 This subsection contains a list of the parameters collected by the TeamQuest Baseline system log probe The following statistics are collected for each workload defined on the system Note These parameters are not collected by TeamQuest Online Parameter Hierarchy Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs ES IO SUPs ES IP SUPs CS IO SUPs CS IP SUPs IO SUPs Core Block SUPs IO Requests Disk I O Requests Disk I O Words Tape I O Requests Tape I O Words Voluntary Wait Resource Wait Active Time System Log Workload Total Total SUPs used to process the job Sum of the IP SUPs ER CC SUPs operating system SUPs and I O SUPs Total IP SUPs Total ER and control card SUPs Total Exec Services I O SUPs Total Exec Services IP SUPs Total control statement I O SUPs Total control statement IP SUPs Total I O SUPs Total core block SUPs Sum of SUPs and main storage u
219. ctor inserts them This first retention period is referred to as the primary retention period If records are not collected at regular intervals such as alarm records then the primary retention period is the only retention period that can be specified If the data is intervalized additional retention periods can be defined At the end of the primary retention period the records are aggregated into records whose interval matches the interval specified for the next retention period It is assumed that retention intervals increase from smallest for the most recent records to largest for the oldest records When the end of the oldest retention interval is reached the records are removed from the table Registration Table data is kept in files on the system where the database resides The table list that can be displayed using tqzip shows the name of the file that contains the data for a given table Table files are completely self contained Thus they can be saved to a lower cost storage medium such as tape and reloaded later for examination In order for TeamQuest Baseline to know about the reloaded table files the table file must be registered with TeamQuest Baseline If you do not save table files to tape you will never need to register a table Since the name of the table being registered is likely to match the name of an existing table a new name must be provided for the registered table When a table file registered in this manner is
220. d gt gt File created TQDATA PROD30 minute gt gt Aggregation Set List gt gt Max gt File gt Name Gran Start Stop Keep Size Source Name gt Rae WS a a A Sesh aa a a a poe a Seeker Ree ee Poe oe Ss oh ee gt 1l minute 00 01 00 00 00 00 23 59 59 08 00 00 gt gt 30 minute 00 30 00 00 00 00 23 59 59 35 days gt gt gt An following Name indicates an expanded aggregation set gt gt End TQASA 6 18 TQ 01011 18 Customizing Aggregation Sets 6 4 2 Removing an Aggregation Set You can use the tqasa program to remove an existing aggregation set from a database When you remove an aggregation set the aggregation set file is deleted from the database To remove an aggregation set the tqasa processor call has the following format tqasa d databasename r aggset where d specifies the database for which you want to remove an aggregation set The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 r specifies the name of the aggregation set to remove The aggset parameter is the name of an existing aggregation set Note An aggregation set whose name is greater than 12 characters or that contains an uppercase letter must be deleted using TeamQuest Performance Software Administration Example The following example rem
221. d Info list box in the Workload Maintenance dialog box This causes the selected workload set and all associated workloads to be deleted To delete the workload set definition from the database you must click OK in the Workload Maintenance dialog box You can cancel the deletion of a workload set either by clicking the No button in the alert box or by clicking Cancel in the Workload Maintenance dialog box see 5 3 1 Note When you delete a workload set the database parameters associated with that workload set are also deleted By deleting parameters from the database all of the data values for these parameters are cleared in each of the aggregation sets in the database For this reason you must ensure that all of the database aggregation set files are loaded onto the system and are accessible whenever you are deleting workload sets All space previously used by deleted parameters is automatically reused as new parameters are created Deletion of workload sets should be a controlled operation and is best performed after you have terminated all probes executing on your system Be careful not to delete any parameters that are being used by a probe that is currently executing on your system TQ 01011 18 Customizing Workloads 5 3 5 Creating a New Workload To create a new workload definition 1 TQ 01011 18 Select a workload set or a workload from the Workload Info list box in the Workload Maintenance dialog box The worklo
222. d user alarm code to tqalm ed iq qual file dsuprobe tab TABLE Table_Name 2200 ownerevent File_Name usertabs Table_Time_Event eld_Name Time roup_Key rimary_Key d_Type_Timestamp me_Date_Subtype Seq_Cons_Method Last_Cons FIELD Field_Name System Group_Key Primary_Key Identifier Field_Type_System String_Type Max_ Max_Size 12 Time_Consolidation Seq_Cons_Method Id_Cons FIELD Field_Name Owner Primary_Key dentifier String_Type Time_Date_Subtype ime_Consolidation Seq_Cons_Method Id_Cons FIELD Field_Name Fixed Files Seq_Cons_Method Last_Cons FIELD Field_Name Fixed Tracks Seq_Cons_Method Last_Cons FIELD Oo eof The following command takes the input file qual file dsuprobe tab and creates the table sys lib tq baseline tqtblprb i f qual file dsuprobe tab TQ 01011 18 9 59 Customizing Data Collection Probes 9 60 The following is the data collection runstream that can be placed in file qual file dsuprobe run RUN TQTBL account project id sym df print cat p own prt 1 brkpt print own prt mfdrpt owner_summary eof brkpt print ed uq own prt f Owner 3 on dsplit split tpf a d 222 top i time system top tchar f 2 Gsi NO OWNER ast eof sys lib tq baseline tqtblprb f own prt t 2200 ownerevent end fin TQ 01011 18 Customizing Data Collection Prob
223. dSpec Count ON if c END TABLESPECS D 11 22 TQ 01011 18 Customizing Tables The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The first line of the table specification file must be the header line BEGIN TABLESPECS It is followed by a series of table specifications B This line begins the first table specification Each table specification begins with the class table name in brackets The number of lines in each set of table specifications is variable The lines must appear in order and the label must appear exactly as shown followed by an equals sign Each of the values on a line must be separated by a comma or a space The lines are as follows class table The class and table name of a table that is registered in the performance database RetentionStart The synchronization time for aggregated records All ageregated records will start on even intervals of the retention start time RetentionStart is specified as hhmm and must be a value greater than or equal to 0000 and less than 2400 If no value is specified for this field 0000 is used as the default A value of 0000 is always used if the table does not contain intervalized data PrimaryRetention The length of time the records created by the data collectors are retained This field is specified as nUNITS n must be an integer greater than zero and UNITS can be
224. data series is graphed and which type of trend line is displayed Series The number of the data series that is graphed The number is determined by its position displayed in the series list in the Trend Graph Setup dialog box The top most series is number 1 Line Specifies the type of trend line Valid trend line types are Linear Straight best fit line Polynomial A trend polynomial curve of any order from 2 through 5 Order is determined by the order value Order The order of the polynomial when using the polynomial line type Valid values are 2 3 4 or 5 Equation The setting to determine if the best fit line equation is displayed The text attributes are defined in the BF Note key Valid values are ON Display the best fit equation OFF Do not display the best fit equation Scatter XSeries YSeries Line Order Equation The specifications for the creation of a scatter graph This determines which data series are graphed and which type of best fit line is displayed XSeries The number of the data series that was used along the x axis The number is determined by its position displayed in the x axis list in the Scatter Plot Setup dialog box The top most series is number 1 YSeries The number of data series that was used along the y axis The number is determined by its position displayed in the y axis list in the Scatter Plot Setup dialog box The top most series is number 1 Line Specifies the type of best fit line
225. ded on a user defined report See Section 8 for more information on derived statistics TQ 01011 18 3 5 Part 2 Customizing Your Site Section 4 TeamQuest Performance Software Administration 4 1 4 2 Introduction TeamQuest Performance Software Administration allows you to perform the maintenance functions for workloads aggregations and derived statistics through a graphical user interface What is a Session Before performing any maintenance tasks using TeamQuest Performance Software Administration a connection must be made to the system where the data is being stored This connection to the data is called a session A session defines how you connect to the host where the performance database exists A session definition identifies the name of the particular performance database you are using Once all this information is obtained it is saved and identified by a single name which can then be used on subsequent queries of that data Each session that is opened requires a separate communications connection The total number of sessions that can be opened at the same time is limited to a maximum of 15 and may be further limited by the available memory on your workstation Multiple session connections to the same host may be restricted by the number of connections available When you start TeamQuest Performance Software Administration you immediately encounter sessions The interface will prompt you to open a new session
226. determines how line cluster bar stacked bar and scatter graphs appear XAxis Determines the values placed along the horizontal axis of the report LegendEntries Determines the source of the data series Usually this field contains parameter Page Determines how to separate and sort data for additional pages of a report The following values can be used for the XAxis LegendEntries and Page fields parameter Date Time and keyname A special keyname of value represents the Value Type field of a parameter group You cannot mix the specifications of parameter and keyname You can concatenate the specifications by using a plus sign Concatenated specifications are sorted by the order that the keywords are defined and then by date and time The values contained in the XAxis LegendEntries and Page fields must not contain the same keys If the date and time specifications are not used in any of the fields they are appended to the end of the Page field If any unused keyname values are present they are appended to the Page field For example in the following Format statement Format Resource Statistic Time Date The data for the x axis is sorted by resource and statistic The legend is grouped according to time The pages of the report are ordered by date Report Definition File Trend Series Line Order Equation The specifications for the creation of a trend graph This determines which
227. dinate of the top left hand corner of the window C 22 TQ 01011 18 Report Definition File bottom Specifies the pixel location for the lower coordinate of the bottom right hand corner of the window right Specifies the pixel location for the right coordinate of the bottom right hand corner of the window Background Color Color The color for the background of the report displayed The default value is obtained from the View Defaults settings of the User Preferences If a color is not specified TeamQuest View uses the default value See the TeamQuest View User Guide for information on User Preferences BF Note Top Left Bottom Right Vertical Horizontal Font Style Size Foreground Background Border The specifications for the text attributes associated with the best fit equation annotation The text appears in a rectangle defined by the relative coordinates with a range of 0 through 10 000 Top Specifies the relative location for upper coordinate of the top left hand corner of the best fit annotation box Left Specifies the relative location for the left coordinate of the top left hand corner of the best fit annotation box Bottom Specifies the relative location for the lower coordinate of the bottom right hand corner of the best fit annotation box Right Specifies the relative location for the right coordinate of the bottom right hand corner of the best fit annotation box Vertical Defines the vertical alig
228. dminstration erena A 4 4 Creating a Session ooe eeeececseeeeeeeeceeeeeeeeeeeeeeseeeeeeeaeeeeaes 4 4 1 Using a Socket Connection essssssessreeseeerses 4 4 2 Using a SSL Connection sessen 4 4 3 Using a Named Pipe Connection essees 4 5 Additional Sessions ooo eeeeeeeeeteeeeeeeeeeeeeeeeeeeeeeeaeeeeeneees 4 6 Session Timeout Warnings cececceesseeeeeeeeeeeeteeeeeees 4 7 Terminating the Communications Session Customizing Workloads 5 1 INMFOCUCTION citsesizrnaseietecteien died TE 5 2 CONCEPTS erpen a cl ee ieee alee 5 2 1 Workload Qualifiers oaeee neeese 5 2 2 Workload Definition Language eeesseseeeesee 5 2 2 1 Using Numerical Conditions c 5 2 2 2 Using String Conditions ou eee 5 2 2 3 Using Regular Expressions in String CONGILIONS soiien n 5 2 2 4 Using Boolean Expressions in Definitions 5 2 2 5 Definition Example ssesssseeessrressrrrrresrsernn 5 3 Maintaining Workloads with TeamQuest Performance Software Administration eecceeceeeceeseeeeeteeeeeneeeees 5 3 1 Performing Workload Administration Tasks 5 3 2 Creating a New Workload Set Definition _ 5 3 3 Displaying or Modifying a Workload Set Definition 5 3 4 Deleting a Workload Set o eesececeeeeeeeeeesteeeteees 5 3 5 Creating a New Workload sessies 5 3 6 Displaying or Modifying a Workload Definition 5 3 7 Importing Workload Sets and Workload Definitions 5 3 8 Exporting Workload Sets
229. e Thick MajorTick MinorlTick MajorGrid MinorGrid XAxis Labeling Major Minor Midnight Point Consolidation Enable Points Legend Enable Location top left bottom right Legend Style Font Style Size Foreground Background Border Threshold Color Normal Exception Threshold Pattern Normal Exception Window Area top left bottom right Background Color Color Auto Multiplier ON OFF Display Markers ON OFF BFNote Top Left Bottom Right Vertical Horizontal Font Style Size Foreground Background Border Note Text Top Left Bottom Right Vertical Horizontal Font Style Size Foreground Background Border Export Data Page Spread Comma Sequence Color Hatch Marker Line TQ 01011 18 C 3 Report Definition File The following descriptions explain the syntax of the general specifications of the report definition Copyright Anything You Want The copyright statement is never saved by TeamQuest View This statement is used by TeamQuest View when it reads this definition to identify that it is not to overwrite the file when you use the File gt Save menu selection You can still overwrite this definition by using the File gt Save As menu selection and specifying the same file name When you save the report this way the copyright statement is not written to the file Product Name text The name of the product for which this report i
230. e as shown in this example After you have selected a table tqzip displays the following Collection Period Menu Collection Period Menu 2200 Runs 1 List Collection Periods 2 Add a Collection Period gt gt gt gt gt 3 Delete a Collection Period gt 4 Turn Data Collection Off gt gt gt gt gt 98 Omit Collection Period Changes 99 Save Changes and Exit Collection Period Menu Enter your selection lt 1 gt To perform a collection period task enter its associated number at the prompt To accept any changes you may have made on subsequent collection period screens and to return to the Tables Menu enter 99 To omit any changes and to return to the Tables Menu enter 98 TQ 01011 18 11 5 Customizing Tables 11 5 1 Listi ng Collection Periods To display a list of the currently defined collection periods for the table select List Collection Periods from the Collection Period Menu by entering 1 at the prompt The tqzip program displays the following list gt gt gt gt gt gt gt gt gt gt gt gt gt Enter your selection lt 1 gt 1 Data Collection is On Collection Periods 2 A Exclude umber Start End SMTWTFS Interval B 00 00 08 00 lt No Collection C 1 08 00 11 00 S S 00 01 00 11 00 13 00 lt No Collection gt 2 13 00 16 00 S S 00 00 30 16 00 24 00 lt No Collection gt Continue The following lettered descriptions corre
231. e 3 1 u specifies that an unload operation is to be performed by tqzip The type of unload operation is derived statistics as indicated by the s option S unloads the derived statistic definitions from the database to the file specified by the derived file parameter 8 18 TQ 01011 18 Customizing Derived Statistics Derived Statistic Definition File Format The derived statistic definitions appear in the following format in the output file created by tqzip If you update the derived statistic definition file you should ensure that the format of the file remains consistent with what was output by the u and s options The following example shows how the definitions appear in the file BEGIN DERIVED STATISTICS unloaded on Fri Jan 13 17 46 30 2012 A i B Statistic Identifier C category group System Log category Workload subcategory Demand statistic Population open runs Informational Text D Population open runs Formula is used as the newline character E active elapsed Aliases F alias name active G used in formula Y H system gt SYSTEM I category group System Log category Workload subcategory Demand statistic Active Time appl group ane workload set gt WLS workload so SW resource ee alias name elapsed J used in formula Y system gt SYSTEM category group System category PN subcategory ot statistic Elapsed Time appl gr
232. e 5 3 8 Deleting a Workload see 5 3 9 TQ 01011 18 5 3 1 Customizing Workloads Performing Workload Administration Tasks To begin any workload maintenance task select Administration gt Workloads If more than one session is open the Administration Session Selection dialog box is displayed This dialog box contains a list of open sessions for which you can perform workload maintenance and session information for the session highlighted in the list box Select the desired session and click OK If more than one database is available for the session the Administration Database Selection dialog box is displayed Select the database for which you want to perform workload maintenance and click OK The Workload Maintenance dialog box is displayed Note As the workload information is loading the status bar at the bottom of the dialog box indicates the percentage of information loaded When the percentage reaches 100 the message disappears and you are ready to perform workload maintenance tasks 4 Workload Maintenance Session cl01 Database PRODUCTION Oo 7 Workload Info Cancel Heo oen New ew workload set Import Export iH Workload information received in Update mode All workload maintenance tasks can be performed through this dialog box The Workload Info list box contains the list of source data types workload sets and workloads defined for the database and session listed
233. e MAPPER log probe Workload Definition Language You use the workload definition language to specify the definitions of your workloads This language is used when specifying workload definitions through TeamQuest View A workload definition is a Boolean combination of conditions A definition is expressed as the combination of a characteristic name an operator and a value This combination is referred to as a condition A condition can be either a numerical condition or a string condition depending on the characteristic used in the definition A numerical condition is used when the values for the characteristic must be numeric A string condition is used when the values for the characteristic must be a string Conditions can be combined using the Boolean operators OR AND and NOT to form complex definitions The following are examples of workload definitions PID 111 Run_ID CMS TQ 01011 18 5 2 2 1 Customizing Workloads Using Numerical Conditions A numerical condition is expressed as a characteristic name a numerical operator and a numerical value set An example of a numerical condition is PID lt 111 In this example PID is the characteristic name lt is the numerical operator and 111 is the numerical value set When creating a numerical condition the following rules apply e The characteristic name can be any of the characteristics associated with a specific probe listed as an integer or real number ty
234. e all records whose timestamp is before a specified time Select Delete Records Before Time from the Data Deletion Menu by entering 1 at the prompt The following prompts are displayed Enter your selection lt 1 gt 1 Delete all records before the following date and time Enter date mmddyyyy A 01132012 Enter time hh mm B gt 00 00 SME MSN NEM N SS The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A Enter the date before which all records will be deleted B Enter the time before which all records will be deleted In the example all records whose timestamp is before midnight on January 13 2012 will be deleted from the table 11 7 2 Deleting Records for a System The tqzip program allows you to delete all records for a given system from the table Select Delete Records for System from the Data Deletion Menu by entering 2 at the prompt The following prompts are displayed gt Enter your selection lt 1 gt gt 2 gt gt Enter system to delete gt venus At the prompt enter the system name for which all records will be deleted In the example all records whose system is venus will be deleted TQ 01011 18 mar Customizing Tables 11 7 3 11 8 11 18 Deleting All Records for a Table The tqzip program allows you to delete all records for a table Select Delete All Records from the Data
235. e and click OK The Derived Statistic Maintenance dialog box is displayed Note As the derived statistic information is loading the status bar at the bottom of the dialog box indicates the percentage of information loaded When the percentage reaches 100 the message disappears and you are ready to perform derived statistic maintenance tasks 4 Derived Statistic Maintenance Session cll Database PRODUCTION Oo ade Cancel Derived Statistics PRODUCTION Help Detived Processor I0 Activity Memory System Log HVSTAT Log Bow Mapper Log Import Export FREE Derived statistic information received in Update mode All derived statistic maintenance tasks can be performed through this dialog box The Derived Statistics list box contains a hierarchical list of the derived statistics that are defined in the database see 8 3 2 When the list initially displays the hierarchy is expanded to the first level that contains more than one derived statistic The database name is at the top followed by the category group category subcategory and statistic TQ 01011 18 8 5 Customizing Derived Statistics 8 3 2 The status bar at the bottom of this dialog box displays information about the selected item in the Derived Statistics list box It displays status information about the derived statistics maintenance tasks or tells you the type of item database name category category group subcategory or statistic t
236. e average time required to process the I O and the average time required to transfer the data SUMMARY RPT Summary Information Gives statistics for each workload on the number of I O requests I O requests that were reads I O requests that were cached number of words per input request number of words per output request the average time required to process the I O and the average time required to transfer the data Predefined Report Descriptions B 5 B 6 B 14 MIPS Report The MIPS report described in this subsection provides information about MIPS usage for a Capacity On Demand COD system The MIPS report is found in the mips subdirectory Note This report is not available for TeamQuest Online File MIPS RPT Name MIPS Usage Description Reports MIPS usage information including average MIPS used Baseline MIPS Ceiling MIPS Desired MIPS and Running MIPS MAPPER Log Reports The MAPPER log reports described in this subsection display statistics for the defined workloads and general statistics for the system The following workload report definitions can be found in the mprlog subdirectory Note These reports are not available for TeamQuest Online File ACTIVE RPT Name Active Time Description Gives statistics for each workload on total function active time in milliseconds File BRKPTS RPT Name Breakpoints Description Gives statistics for each workload on the number of breakpoints that occurr
237. e current system as returned in the program s initial register set is used specifies the category group key name in the parameter hierarchy of the statistic being stored If the category group is not specified User Defined is used for the category group specifies the category key name in the parameter hierarchy of the statistic being stored If the category is not specified null is used for category specifies the subcategory key name in the parameter hierarchy of the statistic being stored If the subcategory is not specified null is used for subcategory specifies the aggregation set to use for storing data The aggset 1 parameter is the name of an aggregation set defined in the database An aggregation set is strictly optional on the command line If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used prints usage information containing command line syntax and exits required keyword which ends command line option scanning 9 11 1 3 Installing Your User Probe After you have finished developing your user probe you may want to edit the control stream in SYS LIB TQ BASELINE TQCTL ADD so that your probe is started along with the other probes 9 11 1 4 Restrictions The following restriction applies to the use of the tqusrprb program You are limited to one Category Group Category SubCategory in the parameter key hierarchy TQ 01011 18 9 47
238. e database the tqzip program displays the following prompt to turn off threshold checking gt Enter your selection gt 5 gt Threshold checking is ON for TQDATA PRODUCTION gt Enter Y to reverse this setting gt Threshold checking is now OFF All active TeamQuest Baseline probes automatically stop threshold checking at the next data sample when Threshold Checking is changed from ON to OFF TQ 01011 18 7 25 Customizing Alarms and Thresholds If threshold checking is currently turned off for the performance database tqzip displays the following prompt to turn on threshold checking gt Enter your selection gt 5 gt gt Threshold checking is OFF for TQDATA PRODUCTION gt Enter Y to reverse this setting gt y gt gt Threshold checking is now ON gt DHAKA KKK KKK KK KK KKK KKKK KKK KKK KK KK KKK KKKKK KKK KK gt TQALM must be executing on your system gt when threshold checking is ON X yhhkxkxk kxkkxkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkxk All active TeamQuest probes automatically begin threshold checking at the next data sample when Threshold Checking is changed from OFF to ON 7 15 2 Using the tqzip Processor Call To turn off or on threshold checking or to view the current setting for threshold checking use the following processor call format tqzip d database a action where d specifies the database that tqzip uses The database parameter is the name of an existing database that
239. e each number sign represents one digit The text is the heading for the column of values The text must be enclosed in double quotation marks If the text specification is not included the column heading defined for the table is used The row col is optional and indicates the row and column on the page where the text string should start The parentheses must be included in the row col specification You can specify the row or column number in one of two ways To indicate the absolute row or column specify only a number for the row and column To indicate a relative position the number should be preceded by a plus sign The relative position is calculated from the position where the previous character was printed on the page Optionally you can indicate just a row specification by row In this case the text is centered on the page If the position of the column is not specified the column is placed next to the previous column from the table If this is the first column the column starts in the left most position on the page You can specify more than one DataFields key Host Based Reporting for Table Data E 3 3 E 10 MaxRecs nnn The maximum number of records to print Any records after this number will be discarded Sort field1 ASCENDING DESCENDING field2 ASCENDING DESCENDING fieldn ASCENDING DESCENDING The sort criteria for the records field1 field2 and fieldn are v
240. e file that tqzip should use for a reduction load or unload operation The reduction file parameter is the name of the file containing the reduction definitions for your database The r option must be used with the l or u option S specifies the file that tqzip should use for a derived statistics load or unload operation The derived file parameter is the name of the file containing the derived statistic definitions for your database The s option must be used with the l or u option See 8 4 1 and 8 4 2 t specifies the file that tqzip should use for a threshold load or unload operation The threshold file parameter is the name of the file containing the threshold definitions for your database The t option must be used with the l or u option See 7 13 and 7 14 W specifies the file that tqzip should use for a workload load or unload operation The workload file parameter is the name of the file containing the workload definitions for your database The w option must be used with the l or u option See 5 4 1 and 5 4 2 H 2 TQ 01011 18 Appendix 2200 Socket Interface TeamQuest Baseline uses the TCP IP interface supplied by the Communications Application Program Interface to provide the capability of connecting from TeamQuest View or TeamQuest Performance Software Administration to tqdata using an OS 2200 system socket To use this connection option you need to 1 TQ 01011 18 Install the Communications Application Prog
241. e following example shows a workload definition using program names to identify system transactions This workload contains information for system transactions CSF0 through CSF99 XYZ and SYSERR By using the Txn_Code CSF 0 9 0 9 construct data for transactions with identifiers of CSFO and CSF9 and CSF00 through CSF99 will be included in this workload Txn_Code CSF 0 9 0 9 XYZ TQ 01011 18 TRSY SERRE 9 23 Customizing Data Collection Probes 9 7 6 Log File User Area This format of the user area is common to some sites using the Logic and Information Network Compiler LINC 1100 The following shows the format of the log file user area Word 0 Word 1 Word 2 Word 3 Word 4 Word 5 where Length Version Status Filler Txn_ Run_ID ISPEC_Name Timestamp Appl _Group Sentinel 9 24 Length Version 0 Status Filler 1 Txn_Run_ID 2 ISPEC _Name first 4 chars 3 Timestamp 4 ISPEC_Name 5 5th char Appl_Group Sentinel LN is the length of the user area currently 6 words 18 bits is the version of the user area 18 bits is the status of record A zero value indicates the record is OK 6 bits is not used 30 bits is the unique transaction run id of this transaction 6 FIELDATA characters is the name of the initially called control program The first four characters are in word 3 with the 5th character in Q1 of word 5 These are
242. e for information on User Preferences Value Style Font Style Size Foreground Background Border The specifications for the text attributes of the title of the vertical axis Font Specifies the typeface in which the value title is printed Style Specifies the kind of emphasis for the typeface of the value title Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the title be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Defines the number of points for the typeface of the value title This value determines the height of the characters in which the value title is printed The value must be a positive integer Foreground Indicates the color in which the value title is printed TQ 01011 18 Report Definition File Background Indicates the color of the area that appears directly behind the value title of the axis The background area is a boxed area and acts as a frame for the value title If a background color is specified the boxed area behind the value title covers any graph lines and other text that may appear behind the value title If the Background field for the value title is not specified TeamQuest View allows any graph lines or other text that may be behind t
243. e or update another threshold definition enter y at the prompt The tqzip program returns you to the create or update threshold prompt in A If you have finished creating threshold definitions enter n at the prompt R When any changes have been made to the threshold definitions in the performance database the tqzip program displays a final prompt If overall threshold checking is turned on you must notify any active probes of the threshold definition changes before they will take effect Enter y at the prompt to notify the probes You can also use the Notify Probes of Threshold Changes selection on the Alarms and Thresholds Menu to perform this task see 7 12 If overall threshold checking is turned off the tqzip program displays the following prompt gt Threshold checking is OFF gt Do you want to turn it ON y lt n gt gt y Threshold checking is now ON yhhkkk kkk kkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkkk gt TQALM must be executing on your system bas gt x when threshold checking is ON K yhhkxkk kkxk kkkkkkkkkkkkxkkkkkkkkkkkkkkkkkkkkkkkkkkkk v Overall threshold checking must be turned on before any of the individual threshold definitions are checked for threshold violations To turn on overall threshold checking you can either answer y to this prompt or use the Turn Threshold Checking On Off selection on the Alarms and Threshold Menu to perform this task When using threshold checking you must also ensure th
244. e parameter_selection definition can be repeated as many times as needed The parameter_selection definition is specified using the parameter name all on a single line with the key names separated by colons The order of the key names is system category group category subcategory statistic resource application group workload set workload The parameter_selection definition is case sensitive All colons must be included even if a parameter key name is blank A single mask character can be specified in each key name field of the parameter selection The mask can be specified as string string or string string where string is any string of characters To see a list of parameters currently defined in the database enter a question mark at the parameter selection prompt You terminate parameter selection prompting by entering a blank transmit at the parameter selection prompt If you enter any type of control statement including EOF as a response to a prompt tqrpt terminates immediately without producing a report TQ 01011 18 D 3 Host Based Reporting for Aggregation Set Data D 1 D 4 Using tqrpt for Condensed Output When the x option is used with tqrpt the report output is in a condensed format with comma or tab delimiters This format is suitable for importing into a spreadsheet program The length of a line in the condensed output is controlled by two factors a maximum field count and a max
245. e predefined reports for your host system Appendix C Report Definition File This appendix describes the format of the report definition file It provides the format of the file and a description of the fields specified in the file Appendix D Host Based Reporting for Aggregation Set Data This appendix describes how to use the TeamQuest Baseline tqrpt program to generate reports containing aggregation set data directly from the performance database on the host system This program is available with TeamQuest Baseline only Appendix E Host Based Reporting for Table Data This appendix describes how to use the TeamQuest Baseline tqrep program to generate reports containing table data directly from the performance database on the host system This program is available with TeamQuest Baseline only Appendix F The tqgetm Program for TeamQuest Model This appendix describes the tqgetm program that allows you to export data from TeamQuest Baseline to TeamQuest Model Appendix G Alarm History Table This appendix describes the format of the alarm history table Appendix H tqzip Processor Call Format This appendix serves as a reference for the tqzip processor call format Appendix I 2200 Socket Interface This appendix provides the information necessary to connect an OS 2200 system to TeamQuest View using a socket connection Appendix J 2200 SNMP Trap Alarms This appendix describes how to define thresholds that will send alarm event
246. e report If Enable is set to OFF the horizontal axis does not appear on the report The default for this field is ON Thick Defines the weight in pixels of the horizontal axis line The value for this specification must be a positive integer The acceptable values are 1 2 and 3 The default value is 1 MajorTick Defines the type of marks used for the major ticks on the horizontal axis line The acceptable values are Invisible No tick marks appear Inside Tick marks appear on the inside of the axis line Outside Tick marks appear on the outside of the axis line Crossing Tick marks cross the axis line The default value for the MajorTick field is crossing Report Definition File C 16 MinorTick Defines the type of marks used for the minor ticks on the horizontal axis line The acceptable values are Invisible No tick marks appear Inside Tick marks appear on the inside of the axis line Outside Tick marks appear on the outside of the axis line Crossing Tick marks cross the axis line The default value for the MinorTick field is invisible MajorGrid Indicates whether grid lines are drawn across the graph at the points of the horizontal axis designated by the major ticks The acceptable values are ON or OFF If MajorGrid is set to ON the grid lines appear on the report If MajorGrid is set to OFF the grid lines do not appear on the report The default value is OFF MinorGrid Indicates whether grid lines are drawn acr
247. e they can be used without affecting the existing production workloads TeamQuest Performance Software Administration and the host based tqzip program are used to perform workload management tasks on your system Workload statistics can be viewed using the TeamQuest View application 2 5 Alarms and Thresholds Note The alarms and thresholds capability is not available with TeamQuest Online The TeamQuest Baseline probes collect data from various data sources at the sample interval defined for the user specified aggregation set The probes can also check the data obtained at each sample interval to determine if the data values exceed user specified threshold limits If a data value for a parameter is found to exceed a threshold the probe can generate an alarm to alert you that a threshold condition has been met The tqalm program handles all alarms generated by the TeamQuest Baseline probes tqalm sends console messages makes the alarm messages available for display by TeamQuest View and performs site specific actions for each alarm The tqalm program can optionally send an SNMP trap when an alarm occurs see Appendix J The tqalm program also performs threshold checking on any derived statistics for which thresholds are defined TQ 01011 18 2 5 TeamQuest Baseline and TeamQuest Online Concepts 2 6 2 6 Derived Statistics Note The derived statistic capability is not available with TeamQuest Online TeamQuest Baseline allows y
248. e tqmpr processor call statement See 9 9 2 for the format of the processor call statement The format for these parameters is as follows f accentg log n L f accentg log nt 1 S system name where f specifies a MAPPER accounting log file to be processed accntg log is the name of the file n is the file cycle number When processing multiple MAPPER accounting log files multiple f options are needed one for each cycle to be processed All cycles must be in sequential order S specifies the system name that identifies the system source of the MAPPER accounting log file data It is used to determine where to store the data in the performance database The system name parameter is case sensitive The tqmpr d a and p parameters see 9 9 2 are also used for MAPPER accounting log file input Only one aggregation set a can be specified when using the f parameter The a and p parameters are required on the tqmpr processor call If you do not want tqmpr to load the MAPPER data into the version of the performance database specified on the Database setting in the tq ini file use the d parameter to specify an alternate database version Note Before loading data into an alternate version of the performance database you must use the tqdbu utility to create this alternate version see 10 4 The tqmpr program processes each MAPPER log file storing data points in the performance database as specified by the aggregation
249. e traps to network manager s that are listening on port 162 on the management host s To send SNMP traps for alarms you need to 1 Update your CMS 1100 configuration to define the TQSSU process and the trap destinations see J 2 2 Create TeamQuest Baseline threshold definitions that contain the SNMPTrap command in one of the user text fields see J 3 3 Update your network manager configuration to allow it to recognize and accept TeamQuest Performance Software SNMP traps see J 5 4 Ensure that threshold checking is turned on in TeamQuest Baseline and that the tqalm run is executing on your system The TeamQuest data collection probes that continuously execute on your system must also be up and running TQ 01011 18 J 1 2200 SNMP Trap Alarms J 2 J 3 CMS 1100 Configuration To use the SNMP TSAM interface for SNMP traps you need to add a PROCESS statement for TQSSU to your CMS 1100 configuration This statement defines a process that allows tqalm to initialize itself as an SNMP service user SSU The following PROCESS statement must be used PROCESS TQSSU TYPE TSAM PASSWORD TEAMQUEST To define the destination systems for your SNMP traps use the SNMP MGMT statement in your CMS 1100 configuration This statement defines the community name and host addresses that are to receive the traps For example SNMP MGMT MANAGE COMMUNITY NAME public WRITE ACCESS NO TRAPS public xXxXX XXX XXX XXX TRAPS public X
250. e type used by most fields This is the default Field_Type_Timestamp is the time of the sample Field_Type_Desired_Interval is the interval of the sample Field_Type_System is the system for this data Consolidation Methods The cons method field specifies the consolidation method for the Seq_Cons_Method and the Non_Seq_Cons_Method keywords in a FIELD definition see Field Definition in 9 11 2 1 The following are allowable values for cons method Id_Cons Use the identifier for consolidation Use if the identifiers are the same Otherwise NULL the value This is the default Min_Cons Use the minimum of the values for consolidation Max_Cons Use the maximum of the values for consolidation Sum_Cons Use the sum of the values for consolidation Weight_Cons Use the weighted summation of the values for consolidation First_Cons Use the first value time ordered for consolidation Last_Cons Use the last value time ordered for consolidation TQ 01011 18 9 65 Customizing Data Collection Probes 9 66 Data Types Definition Each data value must be formatted according to the data type of the field and in some cases according to the data subtype see Field Definition in 9 11 2 1 The following apply to data value formats Short_Int_Type Subtype is not checked for this type Allowable characters are numeric digits the plus sign and hyphen if they precede the numeric digits Any other character will cause the
251. e_Date_Subtype eq_Cons_Method Last_Cons ield_Name Interval roup_Key rimary_Key jeld_Type_Desired_Interval apsed_Time_Subtype eq_Cons_Method Sum_Cons ield_Name System Group_Key Primary_Key Identifier Field_Type_System tring_Type Max_Size 24 Time_Consolidation Seq_Cons_Method Id_Cons FIELD Field_Name Account Primary_Key Identifier String_Type Max_Size 12 ime_Consolidation Seq_Cons_Method Id_Cons FIELD FIELD 1nMMNMVMHOAMMmNaAnVGonTt FIELD nN e TQ 01011 18 9 55 Customizing Data Collection Probes FIELD Field_Name of Fixed Files Seq_Cons_Method Last_Cons FIELD Field_Name Disk Space Seq_Cons_Method Last_Cons EOF The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A This line specifies the table definition The table definition is continued on the next line by using the character B The line specifies the first field definition This field definition is continued on the next five lines by using the character The other field definitions follow the first definition Each definition begins with the work FIELD and starts on a separate line The following command takes the input file TPF dsuprobe tab and creates the table filename tqtblprb i f dsuprobe tab This command only needs to be run once to create the
252. ec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Memory Program Load TIP Mem Trans Abs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load TIP Mem Trans ZOOMs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load Non TIP Mem Trans Abs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second TQ 01011 18 OS 2200 Parameters Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Memory Program Load Non TIP Mem Trans ZOOMs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads pe
253. ec I O processing Percentage of Exec processor control processing Percentage of Exec memory management processing Percentage of Exec TIP integrated recovery and audit control processing Percentage of Exec symbionts processing Percentage of miscellaneous switchable Exec processing Percentage of user real time processing Percentage of independently linked Exec subsystems Percentage of CMS 1100 processing Percentage of UDS processing Percentage of TIP processing TQ 01011 18 OS 2200 Parameters Critical Deadline Demand Batch Category Group Category Resource Subcategory Statistic Name Hardware Default Unretryable Hwr Chk Diagnostic Interrupt 3 Interrupt 4 Interrupt 5 Interrupt 6 Interrupt 7 Reference Violation Addressing Exception Terminal Addr Exception RSC Underflow Overflow Signal Test and Set Invalid Instruction Page Exception Arithmetic Exception Data Exception Operation Trap Breakpoint Quantum Timer Interrupt 21 Interrupt 22 Interrupt 23 Software Break Jump History Full Delayed Hardware Check Dayclock TQ 01011 18 Percentage of critical deadline processing Percentage of demand processing Percentage of batch processing Processor Java Processor IPO IPI Hardware Interrupts Hardware default interrupts per second Unretryable hardware check interrupts per second Diagnostic interrupts per second Interrupt 3 not used interrupts Interrupt 4 not used int
254. ec Sec Lock Blk Sec Lock File Sec Maint Lock Sec Immed Lock Rec Sec Immed Lock BIk Sec Lock String Sec Lock Blk Rec Prb Sec Lock Rec Prb Sec Category Group Category Subcategory Statistic Name Unlock Rec Unlock Blk Unlock Proc IO Activity Multihost UDS RLP Locks Total record locks by UDS Total block locks by UDS Total file locks by UDS Total maintenance locks by UDS Total immediate record locks by UDS Total immediate block locks by UDS Total string locks by UDS Total block record and probe validity locks by UDS Total record and probe validity locks by UDS Total record locks per second by UDS Total block locks per second by UDS Total file locks per second by UDS Total maintenance locks per second by UDS Total immediate record locks per second by UDS Total immediate block locks per second by UDS Total string locks per second by UDS Total block record and probe validity locks per second by UDS Total record and probe validity locks per second by UDS IO Activity Multihost UDS RLP Unlocks Total record unlocks by UDS Total block unlocks by UDS Total process unlocks by UDS TQ 01011 18 OS 2200 Parameters Unlock Subcoll Unlock File Unlock String Release Blk Rec Unlock Rec Sec Unlock Blk Sec Unlock Proc Sec Unlock Subcoll Sec Unlock File Sec Unlock String Sec Release Blk Rec Sec Category Group Category Subcategory Statistic Name RLP Req Switch Port Init Dump Purge Node Do
255. ecify multiple styles for the report status bar text However the styles must be joined by a plus sign For example to specify that the report status bar text be printed in bold and italic typeface you enter bold italic in the Style field of the specification TQ 01011 18 Report Definition File Size Defines the number of points for the typeface of the report status bar text This value determines the height of the characters in which the report status bar text is printed The value must be a positive integer Point Consolidation Enable Points The specifications that determine how the points of the data series are grouped along the horizontal axis In general it determines the number of data series points from the database that are consolidated to form one data point on the graph Sometimes the number of data points in the data series is more than can actually be graphed in a window When this happens TeamQuest View automatically consolidates data points so that the entire time window can be graphed This concept is called point consolidation Enable Determines if automatic point consolidation is used for the report window The acceptable values are ON or OFF ON indicates that automatic point consolidation is in effect OFF indicates manual consolidation is in effect Points Is an integer greater than zero that specifies the number of data points that are consolidated into one point on the graph when manual consolida
256. ed File CONNECT RPT Name Connect Time Description Gives statistics for each workload on the total time in seconds that stations were connected File FUNCS RPT Name MAPPER Functions Description Gives statistics for each workload on the number of run functions display functions logged functions logged functions outside of runs functions processed logged functions under runs and started functions that occurred under runs TQ 01011 18 Predefined Report Descriptions B 7 File Name Description File Name Description File Name Description File Name Description INTERACT RPT MAPPER Interactions Gives statistics for each workload on the number of terminal interactions that occurred IO RPT MAPPER I Os Gives statistics for each workload on the number of input and output operations processed number input and output operations from functions under runs and the number of input operations from functions under runs LINES RPT Lines Processed Gives statistics for each workload on the number of lines read and written from functions under runs RESPONSE RPT MAPPER Response Gives the statistic for each workload on the number of seconds per response and the number of seconds per logged function System Log Reports The system log reports described in this section provide information about the workloads defined for the system The workload report definition files are
257. ed in a particular way you should catalog the files listed above before creating the performance database Note You should ensure that all of the files for your performance database remain loaded on your system at all times If a database is no longer needed be sure to use tqgdbu to delete it 10 2 TQ 01011 18 10 4 10 5 Customizing Performance Databases Creating Additional Databases You can use the tqdbu utility to create additional TeamQuest databases on your system Since a database is automatically created for you during the installation of TeamQuest Baseline or TeamQuest Online you are not required to use tqdbu to create a database Additional databases may be desirable if you want to freeze an existing database and save it for later analysis while continuing to collect current data from your system into an alternate database To do this you need to use tqdbu to create the new database and then modify the Database setting in the tq ini element see 3 1 You may also want to use an alternate database to load data from a specific set of files using the TeamQuest Baseline or TeamQuest Online probes The tqdbu utility creates a database with one aggregation set in it This aggregation set is always named 1 minute and can hold 8 hours worth of data gathered at 1 minute intervals If you want to create additional aggregation sets for the database you can use TeamQuest Performance Software Administration see 6 3 To create
258. ed times or timestamps Selected fields in a table can be used to help identify records in the table These fields are called identifiers Examples of identifiers are system name severity and alarm identifier Tables are created by the TeamQuest Baseline probes Table records are inserted by TeamQuest Baseline and the TeamQuest Baseline probes as is appropriate for the type of records being created Data Collection TeamQuest Baseline and TeamQuest Online use probes to gather data The probes collect data about the host system on which they are running They also extract statistical values and date and time information for performance parameters and store that information in the performance database Workload Management Note The workload capability is not available with TeamQuest Online TeamQuest Baseline allows you to track the activity on your system based on a logical classification of the work performed on the system This logical classification of work is called workload characterization Workload characterization is the process whereby groups of related performance data are organized into meaningful units called workloads You define workloads for your performance database and the TeamQuest Baseline probes distribute the data they collect according to the defined workloads The workload data is stored in the aggregation sets of the performance database Workload characterization is used to summarize a large set of data in ways th
259. eeeeeteeeeeeeneeeees Using tgal ste aii allio cane A a Alarm Messages oie eeeeeeeceeeeeeeeeeeeeeeeeeeeeeeeeeneeeeeeeneeaaees Site Specific Alarm Handling cccecseeeeeeeeeeesteeees Single Point Operations SPO Considerations _ Customizing Derived Statistics 8 1 8 2 8 2 1 8 2 2 8 2 3 8 3 8 3 1 8 3 2 8 3 3 8 3 3 1 8 3 3 2 8 3 3 3 8 3 4 8 3 5 8 3 6 8 3 7 8 3 8 8 3 9 8 4 8 4 1 8 4 2 Introduction oerrinne latin eels GONMCOPIS Frssisccsivacceestsces tecesseeebe Ti ttehadibe anA aiani Derived Statistic Identification eee Parameter References and Substitution Formula Definition eeeeeceeeeeeeeeeeceeeeeeeeeeeeeees Maintaining Derived Statistics with TeamQuest Performance Software Administration 06 Performing Derived Statistic Administration Tasks Derived Statistic Hierarchy ceeeeeeeeeeeees Creating a New Derived Statistic Definition Defining an Alias oo eee eeeeeeeeeenteeeeeeenaees Qualifying Parameter References _ Deleting an Alias oo eeeeeeeeeeeeenteeeeeeeeaaes Saving a Derived Statistic Definition Viewing Long Key Names eeeeeeteeeeeteeeeeeeeees Importing Derived Statistic Definitions Exporting Derived Statistic Definitions Displaying or Modifying a Derived Statistic DOTININOM a vecesc td ialaiiiel iini ene ae Deleting a Derived Statistic Definition
260. eeeeseeeeseeeeeeeees Adding a Collection Period ccceceeesseeeeeeenees Deleting a Collection Period ccceecesseeeeeeeeee Turning Data Collection On and Off ou Exiting the Collection Period Menu 00008 Defining Retention Specifications eeeeteeeeeee Listing Retention Specifications ou Modifying the Primary Retention Specification Adding a Retention Specification 00 Deleting a Retention Specification Setting the Retention Start Time ou Exiting the Retention Specification Menu _ Deleting Records from Tables oo eeeeceeeeeeeessteeeeeeeenaes Deleting Records Based on Time 0 0 ceee Deleting Records fora System essee Deleting All Records for a Table 2 000 Registering Table perse are a Deregistering a Table eee eeeeeteeeeeeeenneeeeeeeenaeeeeeeeeaas Unloading Table Specifications sseesseeeeeeeeeee ee Loading Table Specifications ececcecseeeeeeteeeeeteeeeee Reorganizing Table Files asssssessssisssesrrrsssrrrressrrennes Why Do I Need To Reorganize My Table Files Executing tarr anec 10 1 10 1 10 2 10 3 10 3 10 4 10 4 10 4 10 6 10 7 10 8 14 1 14 1 14 1 11 2 11 2 11 2 11 3 11 4 11 4 11 6 11 7 11 8 11 9 11 9 11 9 11 11 11 12 11 13 11 14 11 15 11 15 11 16 11 17 11 17 11 18 11 18 11 20 11 21 11 25 11 26 11 27 11 28 TQ 01011 18 Contents Appendix A OS 2200
261. egation sets you can perform the following administrative tasks e Create aggregation sets e Remove aggregation sets e List aggregation sets e Unload aggregation set definitions e Load aggregation set definitions e Convert expanded aggregation sets e Perform data aggregation Concepts The performance database consists of data records saved in various aggregation sets With TeamQuest Baseline you can define additional aggregation sets for your database There are some basic concepts essential to the understanding of TeamQuest Baseline aggregation sets Before defining additional aggregation sets you will want to know about the aggregation set structure the moving windows concept for data storage how aggregation sets are consolidated feeding aggregation sets with data from other aggregation and compact expanded aggregation sets TQ 01011 18 6 1 Customizing Aggregation Sets 6 2 1 Aggregation Set Structure Granularity date grouping retention period and start and end time are concepts essential to the understanding of TeamQuest Baseline aggregation sets To define an aggregation set you specify the following information e The granularity of the data records within the aggregation set Specify the time interval represented by each data point maintained in the database for the aggregation set This interval is called the granularity For example you can specify that TeamQuest Baseline collect and store data samples at
262. egory Statistic Name Demand Resp Time Demand Resp lt 0 5 Sec Demand Resp lt 1 Sec Demand Resp lt 3 Sec Demand Resp lt 5 Sec Demand Resp lt 10 Sec Batch Resp Time Category Group Statistic Name Elapsed Time Active IPs Active Java IPs ES SUP Secs ES SUP TIP Secs TQ 01011 18 System Population Total number of open runs Average number of demand runs open Average number of batch runs open Average number of demand runs in core Average number of batch runs in core Average number of backlogged runs Average number of TIP transactions scheduled System Response Time Average demand response time in seconds Percent of demand responses less than 0 5 seconds Percent of demand responses less than 1 second Percent of demand responses less than 3 seconds Percent of demand responses less than 5 seconds Percent of demand responses less than 10 seconds Average batch response time in seconds System Time in seconds for the sample Current number of non Java Instruction Processors that are active Current number of Java Instruction Processors that are active Exec service SUP charges by non TIP programs in seconds Exec service SUP charges by TIP transactions and batch connect programs in seconds OS 2200 Parameters CS SUP Secs CS SUP TIP Secs IP SUP Secs IP SUP TIP Secs USER IO SUP Secs USER IO SUP TIP Secs Category Group Category Statistic Name T
263. eline maintains workload performance derived statistics that use data from the SIP tqbsp MAPPER log tqmpr and system log tqsys probes The workload performance reports reference these statistics Before using the workload performance reports you must first ensure these derived statistic definitions are in your database You can load the required derived statistic definitions using the following command tqzip l e s sys lib tq baseline derived wpr The following derived statistics are provided e SIP Probe Derived Statistics e MAPPER Log Probe Derived Statistics e System Log Probe Derived Statistics A 8 1 SIP Probe Derived Statistics The following statistics are collected by the SIP tqbsp probe Parameter Hierarchy Category Group Processor Category Total Subcategory Utilization Statistic Name All Overall processor utilization Category Group Processor Category Total Scaled Subcategory Utilization Statistic Name Idle Scaled percentage of time idle Total Exec Scaled percentage of time in Exec processing Total User Scaled percentage of time in user processing Misc NonSwitchable Scaled percentage of miscellaneous nonswitchable IO Processing Scaled percentage of Exec I O processing Processor Control Scaled percentage of Exec processor control processing Memory Management Scaled percentage of Exec memory management processing TIP Integ Rec Scaled percentage of Exec TIP integrated recovery and Aud Cntrl a
264. en multiple of the minimum collection interval defined by the table creator To add a collection period select Add a Collection Period from the Collection Period Menu see 11 5 by entering 2 at the prompt The tqzip program displays the following series of prompts gt Enter your selection lt 1 gt gt 2 gt gt Interval mm ss A gt 1 00 gt Start Time hh mm B gt 8 00 gt End Time hh mm C gt 11 00 gt Mask for days of the week to exclude Sun Mon Sat D gt 1000001 The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A TQ 01011 18 Enter the interval at which data is collected for the collection period If the table does not collect data at regular intervals such as the Alarm Log table then this prompt does not appear Enter the start time for the collection period Enter the end time for the collection period Enter the mask for the days of the week when data is not collected The mask is a series of seven digits corresponding to the days of the week beginning with Sunday and ending with Saturday The mask is specified using the digit 1 to exclude a day and the digit 0 to include a day In the example the mask 1000001 indicates that Sunday and Saturday are excluded from data collection If you do not enter a mask all days of the week are included in data collection Customizing Tabl
265. ent of all requests IO Activity Record Lock Processor Total lock requests per second Total unlock requests per second Total delayed response messages per second Total deadlocks per second Total lock requests Total unlock requests Total delayed response messages Total deadlocks IO Activity Memory File Summary Memory file requests per second Transfer time in milliseconds for memory files File lock delay time in milliseconds for memory files Request existence time in milliseconds for memory files Percentage of time that memory file data was transferred Number of K words per second transferred for memory files Number of K words per I O request for memory files Number of K words per read I O request for memory files Number of K words per write I O request for memory files Percentage of all memory file accesses that were read requests TQ 01011 18 OS 2200 Parameters The following statistics are created if the Exec is level SB4R8 and you have an Extended Processing Complex XPC with the Exec ETCEXEC feature installed Parameter Hierarchy Category Group Category Subcategory Statistic Name Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec Deadlocks Sec Lock Reqs Unlock Reqs Delayed Resp Msg Deadlock Category Group Category Subcategory Statistic Name Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec Deadlocks Sec Lock Reqs Unlock Reqs Delayed Resp Msg Deadlocks Category
266. ent of the product installation file usually SYS LIB TQ BASELINE or SYS LIB TQ ONLINE The default setting is PRODUCTION See 10 4 for information on creating additional databases You must activate data collection before you can use the TeamQuest View software to monitor or analyze your system activity tqctl Runstream A control runstream is provided in SYS LIB RUN TQCTL You can start each probe separately or you can use this runstream to start the data collection probes For TeamQuest Baseline the tqctl runstream also starts the tqalm alarm handler program and the tqagg data aggregation utility Before starting tqetl be sure to terminate all of the runs that it will start that is tqalm tqagg and the data collection runs The tqcetl runstream should be automatically started or scheduled after each system reboot Data will not be collected if the probes are not allowed to run as required on your system TQ 01011 18 9 1 Customizing Data Collection Probes 9 3 9 3 1 SIP Data Collection TeamQuest Baseline and TeamQuest Online contain a SIP data collection runstream called tqbsp The tqbsp runstream uses a run identifier of tqbsp and executes the tqbsp program The tqbsp program collects SIP data and stores it in the performance database The data collected by tqbsp is also available for real time system monitoring The tqbsp program uses the ER SYSBAL interface for collecting SIP data from the OS 2200 system You sh
267. erform the following administrative tasks e List tables see 11 4 e Set collection periods see 11 5 e Define retention specifications see 11 6 e Remove records from tables see 11 7 e Register tables in the database see 11 8 e Deregister tables in the database see 11 9 e Unload data collection and retention specifications see 11 10 e Load data collection and retention specifications see 11 11 e Reorganizing table file in the database see 11 12 11 2 Concepts There are some basic concepts essential to the understanding of the data stored in TeamQuest Baseline tables Before beginning any administrative tasks you will want to familiarize yourself with these concepts 11 2 1 Collections Collections specify the time during a day that data records should be collected for a table If the data stored in the table is collected at regular intervals the interval between collections is also specified The minimum interval allowed is specified by the table creator Any number of collections can be defined for a given table Collection periods cannot overlap each other TQ 01011 18 11 1 Customizing Tables 11 2 2 11 2 3 11 2 4 11 2 Retentions Retentions control how long data is retained in a table If the records are created at regular intervals the retention specifies how long the data should be retained at a given interval Records are kept as they are initially stored at the interval the data colle
268. erformed for TQ Baseline or TQ Online you are asked if you prefer unique user ids for socket connections TQ 01011 18 2200 Socket Interface 1 2 3 1 Same User id For All TQTCP Runs The user id password and account configuration statements are required All other configuration statements are optional and can be used when the defaults are not adequate The following configuration statements are allowed interface interface type This tells TQTCP the communications interface to use when listening for incoming connections The interface type can be either CMS1100 or CPCOMM If this statement is not supplied the default is CPCOMM port port number This tells TQTCP the port number to listen on for incoming connections If this statement is not supplied the default port number is 2700 userid xxxxx password xxxxx The user id and password values are used to initiate the RSI run that is created to handle a socket connection These statements are required and standard OS 2200 rules must be followed for each xxxxx value The user id must have the security attribute specifying that the RUN image is always user entered or RUN image is user entered or system generated Also terminal timeout must be disabled for the user id If you need to use an encrypted user id and password use the euser and epswd statements euser xxxxx epswd xxxxx An encrypted userid euser and password epswd that can be substituted for the userid and pas
269. erived statistic definitions 8 18 threshold definitions 7 21 workload definitions 5 19 updating derived statistic threshold definitions 7 16 threshold definitions 7 12 viewing alarm messages 7 27 derived statistic thresholds 7 18 thresholds 7 13 tq ini element 3 1 TQ 01011 18 Transactions Table System Activity Monitor A 59 turning on and off data collection 11 9 U unloading derived statistic definitions 8 18 threshold definitions 7 21 workload definitions 5 19 updating derived statistic threshold definitions 7 16 threshold definitions 7 12 user data probes tqtblprb for collecting table data 9 38 tqusrprb for collecting performance data 9 38 writing a user probe for performance data 9 39 writing a user probe for table data 9 48 user interface description 1 11 user id encryption I 10 using encryption I 1 V viewing alarm messages 7 27 derived statistic thresholds 7 18 long key names 8 14 thresholds 7 13 W warning screen 4 7 workload definition language 5 2 workload maintenance creating workload definition 5 13 workload set definition 5 10 deleting workload definition 5 18 workload set definition 5 12 displaying workload definition 5 16 workload set definition 5 11 exporting workload definition 5 17 Index 9 Index importing workload definition 5 17 modifying workload definition 5 16 workload set definition 5 11 workload performance derived statistics A 51 workload performance reports B
270. errupts Interrupt 5 not used interrupts Interrupt 6 not used interrupts Interrupt 7 not used interrupts Reference violation interrupts per second Addressing interrupts Terminal addressing exception interrupts per second RCS generic stack underflow overflow interrupts per second Signal interrupts per second Test and set interrupts per second Invalid instruction interrupts per second Page exception interrupts per second Arithmetic exception interrupts per second Data exception interrupts per second Operation trap interrupts per second Breakpoint interrupts per second Quantum timer interrupts per second Interrupt 21 not used interrupts per second Interrupt 22 not used interrupts per second Interrupt 23 not used interrupts per second Software break interrupts per second Jump history full interrupts per second Delayed hardware check interrupts per second Dayclock interrupts per second A 17 OS 2200 Parameters A 2 3 A 2 3 1 A 18 Performance Monitoring Performance monitoring interrupts per second Initial Program Load UPI Initial UPI Normal Initial program load interrupts per second UPI initial interrupts per second UPI normal Memory Statistics The Memory Category Group provides general summary program load and paging statistics The memory utilization statistics are not available from the Exec on OS 2200 on EM64T Server systems As the statistics will exist in the SIP data file wi
271. ers are also required for log file input Only one a aggregation set parameter can be specified when using the f parameter If you do not want tqmips to load the log data into the version of the performance database specified on the Database setting in the tq ini file use the d parameter to specify an alternate database version Note Before loading data into an alternate version of the performance database you must use the tqdbu utility to create this alternate version see 10 4 If the b and e parameters are used with the f parameter tqmips processes the information in the specified log files for the date range specified on the b and e parameters The tqmips program processes each log file storing data points in the performance database as specified by the a aggregation set parameter TQ 01011 18 9 13 Customizing Data Collection Probes 9 6 9 6 1 9 14 Open Distributed Transaction Processing Collection Note The Open Distributed Transaction Processing tqodtp probe is released with TeamQuest Baseline only It is not available with TeamQuest Online TeamQuest Baseline contains an Open Distributed Transaction Processing ODTP monitor runstream called tqodtp The tqodtp runstream uses a run identifier of tqodtp and executes the tqodtp program The tqodtp program collects the most up to date information available on an OS 2200 open standard compliant system environment Statistics on overall system system clien
272. es 11 5 3 Deleting a Collection Period To delete a collection period for a table select Delete a Collection Period from the Collection Period Menu see 11 5 by entering 3 at the prompt The tqzip program displays the following prompts gt Enter your selection lt 1 gt gt 3 gt gt Collection Number to Delete gt gt Data Collection is On gt gt gt Number Start WEN NEON gt gt Collection gt 2 1 2 24 00 lt cancel gt Collection Periods 2 S Number to Delete Exclude SMTWTFS S S Interval lt No Collection gt 00 01 00 lt No Collection gt 00 00 30 lt No Collection gt lt cancel gt The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The tqzip program prompts you for the number of the collection period that you want to delete If you do not know the number of the collection period enter a question mark to list the currently defined collection periods for the table The tqzip program displays the list of collection periods for the table The tqzip program prompts you again for the number of the collection period you want to delete In this example collection number 2 13 00 16 00 is deleted TQ 01011 18 11 5 4 11 5 5 11 6 Customizing Tables Turning Data Collection On and Off The Collection Pe
273. es The following is the user code updated in USRALARM C KKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKKK The process_alarm function is called each time an alarm has been received for processing KKEKKKKKKKKKKKKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KKK KK KKK KKK KKKKKKKEKEKRK ifdef STDC__ static void process_alarm ALARMENTRY alarmentry jfelse static void process_alarm alarmentry ALARMENTRY xalarmentry dfendi f if PRINT_ALARM print_this_alarmentry stdout alarmentry dfendif add user specific alarm processing code here if strcmp alarmentry gt catgroup MassStor by System 0 if strcemp alarmentry gt cat LOCAL 0 if strcmp alarmentry gt subcat 0 if strcmp alarmentry gt stat In Use 0 if alarmentry gt current_zone alarmentry gt normal_zone facsf start qual file dsuprobe run TQ 01011 18 9 61 Customizing Data Collection Probes 9 11 2 4 Command Syntax 9 62 The command line syntax of tqtblprb is filename tqtblprb h t class name i r L f file L d databasename s system end La e You will use the command line options in a particular combination depending on whether you are executing tqtblprb for creating a table inserting data into the table or removing a table from the database Each of the options are described in further detail later in this section If you specify
274. et parameter a You can also set a relative time option when selecting the TPM log entries to process The format of the relative time option is t relative hour where t specifies a relative hour when selecting TPM log entries to process The valid range is 1 to 24 For example t 8 will select and process the last 8 hours of TPM log data relative to when the tqtpm program is started TQ 01011 18 9 7 5 Customizing Data Collection Probes Workload Characteristics for the RTPM and TPM Log File Probes The RTPM and TPM log probes have a unique set of characteristics that are used for its workload definitions You use these characteristics when you create the workload definitions using TeamQuest Performance Software Administration see 5 3 The following characteristics define the RTPM and TPM log workloads Characteristic Valid Type and Valid Name Characters Max Length Operations Txn_Code A Z 0 9 string 6 l PID 0 9 integer 9 lt gt lt gt Program_Name A Z 0 9 string 12 Run_ID A Z 0 9 string 6 Lib_Bank 0 9 string 12 If your site is gathering LINC information in the log file user area you can use the following additional characteristics Characteristic Valid Name Characters ISPEC_Name A Z 0 9 Txn_Run_ID A Z 0 9 Appl _Group 0 9 Example Type and Valid Max Length Operations string 5 string 6 ele integer 2 lt gt lt gt Th
275. ets that follow the rules for feeding aggregation sets For example the granularity of the source aggregation set must be less than or equal to and an even multiple of the granularity of the aggregation set you are defining The date consolidation must also be the same including the first day of the week and the days to exclude Day can feed week and month Week cannot feed month Month cannot feed week See 6 2 4 for more information on feeding aggregation sets Note When you create an aggregation set that is fed by another aggregation set you must execute the tqagg utility program so that data is retrieved from the source aggregation set consolidated and stored in the fed aggregation set See the 6 4 7 for information on the tqagg utility Retain Data For specifies the length of time that data for this aggregation set should be retained in the database Select retention time for the new aggregation set from the options in this list box The maximum retention period that can be selected is dependent upon the granularity of the aggregation set being defined based on a maximum of 31 680 data points that can be stored in one day for any aggregation set Some common granularities and their maximum retention periods that can be selected are Granularity Maximum Retention Period 15 seconds 5 days 30 seconds 11 days 1 minute 22 days 5 minutes 110 days 10 minutes 220 days 30 minutes 660 days 1 hour 1320 days TQ 01011 18 6 11 Custom
276. event classes that were added to the rule base in step 5 The Event Server can be stopped and restarted from the TME desktop GUI or you can use the wstopesvr and wstartesvr commands J8 TQ 01011 18 2200 SNMP Trap Alarms J 5 3 J 6 Computer Associates Unicenter TNG The TeamQuest Performance Software can also send SNMP traps for alarms to the Computer Associates Unicenter TNG Event Console However no predefined message definitions or automated actions are provided as part of this release TeamQuest MIB Definition The following table lists the TeamQuest MIB containing the object identifiers for the enterprise specific variables in the alarm trap This definition is also contained in SYS LIB TQ BASELINE TQMIB TXT The entries in this element can be loaded into a management console to provide textual names for the trap variables if desired Copyright c 2012 TeamQuest Corporation All Rights Reserved No use of this software is permitted except in accordance with the governing license agreement and applicable copyright patent and trade secret laws This file contains the ASN 1 source describing the TeamQuest Corporation enterprise MIB for 0S2200 The enterprise MIB described is subject to change in a future release and support for any object described in this MIB may be removed in a future release as standard MIBs in the management domain described by this MIB become defined and published as
277. f the formula had no data for the time period selected Any required data consolidation is done on each of the components of the formula before solving the formula TQ 01011 18 Section 3 TeamQuest Baseline Administration 3 1 tq ini Element TeamQuest Baseline TeamQuest Online TeamQuest PMLog and TeamQuest Probes use a tq ini element to obtain execution defaults The tq ini element is contained in the installation file for each of the products These files are TeamQuest Baseline SYS LIB TQ BASELINE TeamQuest Online SYS LIB TQ ONLINE TeamQuest PMLog SYS LIB TQ PMLOG TeamQuest Probes SYS LIB TQ PROBES The General Database and Aggregation Sets sections of the tq ini element contain settings that can be updated as needed Do not update any of the settings in the other sections of the tq ini element When updating the tq ini element be sure to update it in all of the product installation files that apply to your site Example tq ini element The following shows a sample tq ini element TQ 01011 18 General Pete Sherine an cs Start of Unix specific settings Log Directory specifies the directory where the log files should be stored A relative or absolute path name can be given Absolute path names must start with a Relative path names are relative to the parent directory of the executable Be sure there is no white space at the end of the line Log Directory log Optional Probes specifies
278. f you want to receive traps from these different types of systems at one network manager you will need to make sure that your network manager configuration properly describes the OS 2200 system traps and the UNIX and Microsoft Windows systems traps Using the network manager configuration elements supplied with the OS 2200 TeamQuest Performance Software and the configuration files supplied with the UNIX Linux and Microsoft Windows systems TeamQuest Performance Software you will be sure to have all of the SNMP trap formats properly defined to your network manager By default tqalm initializes and terminates the TQSSU process in CMS 1100 each time a trap is sent This results in the CMS 1100 slot being allocated for the shortest possible amount of time but also causes the following console messages to be displayed by CMS 1100 each time a trap is sent PROCESS TQSSU IS UP PROCESS TQSSU IS DOWN If you would like to eliminate these console messages update the SYS LIB RUN TQALM runstream to use the i option on the tqalm processor call statement TQ SYS TQALM c i When the i option is present tqalm initializes the TQSSU process once at the start of execution The TQSSU process will not be terminated until tqalm terminates This eliminates the extra console messages but ties up the CMS 1100 slot for the entire time that tqalm is running TQ 01011 18 Glossary A aggregation set A collection of parameter values with each va
279. first need to determine the system and resource that is IP names in which you are interested To report processor busy on IP1 and IP2 for system venus you would enter the following two lines at the tqrpt parameter_selection prompt venus Derived Derived Category Derived Subcategory processor busy IPl venus Derived Derived Category Derived Subcategory processor busy IP2 When each of the derived statistics is calculated venus is substituted as the system name in the alias IP1 and IP2 are substituted as the resource name in the alias Rules for Parameter_Selection Definition The rules for the parameter_selection definition are the same as with stored parameters All keys must be specified in order Keys are case sensitive All colons must be included even if a key name is blank You must be sure to specify the key name for substitution in any derived statistic formula aliases As with stored parameters you can use the mask character to specify any of the keys For example the following parameter_selection would generate a report of processor busy for each IP on your system venus Derived Derived Category Derived Subcategory processor busy TQ 01011 18 Appendix E Host Based Reporting for Table Data E 1 Overview of tqrep TeamQuest Baseline provides the capability to generate reports using the table records stored in your performance database The tqrep program can generate a report based on a user defined report def
280. for reporting from the 30 minute aggregation set on May 8 from 12 00 to 14 00 system category group category subcategory statistic resource appgrp workload set workload sysa Processor Total Utilization Batch sysa Processor Total Utilization Demand sysa 10 Activity Disk Summary Percent Busy DSKF1 sysa System Log Workload Batch I0 SUPs primary OTHER TQ 01011 18 Host Based Reporting for Aggregation Set Data The report would appear as follows System sysa sysa sSysa sysa Category Group Processor Processor I0O Activity System Log Category Total Total Disk Workload Subcategory Utilization Utilization Summary Batch Statistic Batch Demand Percent Busy IO SUPs Resource DSKF1 APPOrp Workload Set primary Workload OTHER Date Time 01 13 2012 12 00 00 45 00 10 00 20 00 15 00 01 13 2012 12 30 00 40 00 10 00 15 00 10 00 01 13 2012 13 00 00 50 00 10 00 12 00 10 00 01 13 2012 13 30 00 45 00 10 00 20 00 17 00 01 13 2012 14 00 00 40 00 10 00 10 00 5 00 D 2 Using tqrpt for Reporting Derived Statistics You can use tqrpt to report the values of derived statistics you have defined in your database Derived statistic definitions are saved in the database as formulas The actual data values reported for a derived statistic are not stored in the database but are calculated at report time You use the p option of tqrpt to generate a list of the stored parameters and derived statistics
281. for this parameter must be in a new zone before an alarm is raised This value should be a multiple of the sample rate Specify how long you want threshold checking disabled for this parameter after an alarm has been raised for it This value should be a multiple of the sample rate This value will not raise an additional alarm An alarm will only be raised when a threshold level is crossed as the value goes up or down Enter a text string that identifies the parameter for which the threshold is being defined This string can be 35 characters or letters in length This string is used for creating the alarm messages that are displayed at the system console Enter the text strings providing site specific information that are displayed when an alarm occurs for this parameter You can enter up to three text strings Each text string must be less than 61 characters in length due to console limitations These lines are optional and if present are added as additional lines in the alarm message sent to the system console Customizing Alarms and Thresholds P The tqzip program displays the threshold definition for approval before it is actually written to the database Then it asks you if you want to save the new threshold definition If you enter y at this prompt the threshold definition is saved in the performance database If you enter n at the prompt the threshold definition is not saved in the performance database Q If you want to creat
282. found in the syslog subdirectory Note File Name Description File Name Description File Name Description TQ 01011 18 These reports are not available for TeamQuest Online TOTSUPS RPT Total SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on total SUPs IPSUPS RPT IP SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on IP SUPs IOSUPS RPT IO SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on I O SUPs Predefined Report Descriptions B 16 File Name Description File Name Description File Name Description File Name Description File Name Description ERCCSUPS RPT ER CC SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on ER and control card SUPs ESCSSUPS RPT ES GCS SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on the Exec Services and control statement IO and IP SUPs CBSUPS RPT Core Block SUPs Gives statistics for the selected workload and environment total TIP demand batch or deadline on core block SUPs SUPS RPT SUPs by Workload Gives statistics for the selected workload and environment total TIP demand batch or deadline
283. g 12 Account A Z 0 9 period string 12 Type a z string 10 The allowable values for Type in the Runs table are demand demand demand rt demand real time batch batch batch rt batch real time tip TIP The allowable values for Type in the Transactions table are self d self destructive self i self initializing reent reentrant onl bat online batch hv tip high volume TIP TQ 01011 18 9 7 Customizing Data Collection Probes 9 4 5 9 4 5 1 9 4 5 2 tqsam Table Data The tqsam probe collects and stores data in two tables the Runs table and optionally the Transactions table The Runs table contains data collected on behalf of demand and batch run activity When the t option is specified on the tqsam processor call line the Transactions table contains data collected on the longest running transactions that are active Information is updated in these tables at the interval rate specified for each table Runs Table Information The Runs table contains information on currently active system demand and batch run activity This information is reported in a tabular column format and is displayed by TeamQuest View See A 9 1 for a list of the information stored in the Runs table by tqsam Transactions Table Information The Transactions table contains information on the top 20 longest running and active transactions By viewing the longest running transactions you can identify those tr
284. g n l f tpmlog n l s system name where f specifies the names of the TPM log files to be processed tpmlog is the name of the TPM log file n is the file cycle number When processing multiple TPM log files multiple f options are needed one for each cycle to be processed All files must be in sequential order S specifies a system name The system name parameter specifies the name of the system source of the system log files It determines where to store the data in the performance database The system name parameter is case sensitive TQ 01011 18 9 21 Customizing Data Collection Probes 9 22 The tqtpm d and a parameters are also required for TPM log file input see 9 7 3 Only one aggregation set a can be specified when using the f parameter If you do not want tqtpm to load the system log data into the version of the performance database specified on the Database setting in the tq ini file use the d parameter to specify an alternate database version Note Before loading data into an alternate version of the performance database you must use the tqdbu utility to create this alternate version see 10 4 If the b and e parameters are used with the f parameter tqtpm processes the information in the specified log files for the date range specified on the b and e parameters The tqtpm program processes each system log file storing data points in the performance database as specified by the aggregation s
285. g the complete parameter hierarchy for the statistic To fully qualify the statistic you would add the system resource workload set workload and application group to the derived statistic hierarchy of category group category subcategory and statistic name see 8 3 3 2 View button allows you to view all key name fields for a reference name in the alias list A key name can be too long to fit in the Alias List edit box and is then truncated The View button allows you to see the entire key name hierarchy see 8 3 5 Delete button allows you to delete a reference name from the Alias List edit box see 8 3 3 3 When you want to use a reference name in a derived statistic formula select the item from the Alias List edit box and click the gt gt button to move the item to the Formula edit box You can also double click the item to move it Function contains a list of server supported functions that can be used in the formula of a derived statistic Select the function and click the gt gt button to move the function to the Formula edit box You can also double click the function to move it TQ 01011 18 8 9 Customizing Derived Statistics Formula provides the mathematical expression that is used to calculate the value for the derived statistic Formulas can consist of parameter references by using aliases arithmetic operators numeric constants and mathematical functions Aliases are used in a formula to reference paramete
286. g trail number Processing from the COD audit trail is much more efficient than using the n option System Log File Input Note Specific COD audit trails cannot be listed for processing since the COD entries within the COD audit trails do not contain local timestamps necessary to store into the TeamQuest performance database As a result only system log files can be input for processing as described here To load data directly from specific log files you must specify two additional parameters on the tqmips processor call statement See 9 5 1 for the format of the processor call statement The format for these parameters is f logfile n f logfile ntl S system name n audit trail number TQ 01011 18 Customizing Data Collection Probes where f specifies the names of the system log files to be processed logfile is the name of the system log file n is the file cycle number When processing multiple log files multiple f options are needed one for each cycle to be processed All files must be in sequential order S specifies a system name The system name parameter specifies the name of the system source of the system log files It determines where to store the data in the performance database The system name parameter is case sensitive n specifies the audit trail number of the system log n 10 unless the system log is locally configured to a different audit trail number The tqmips d and a paramet
287. ganization consolidates this fragmented space When the tqagg program is run it removes records from a table based on the retention information for that table If data is collected for a long period of time without tqagg being run the table file will grow in size If the tqgagg program is then run the records are removed and the space used by the removed records is marked as available for reuse This space is very likely to be in the front or middle of the file For this reason the actual size of the file is not reduced If tqagg is run on a regular basis from that point on the free space in the file is reused but the file size remains constant Executing the tqrorg utility causes the records to be rewritten at the front of the file and any remaining file space to be freed This will decrease the total size of the file As records are added and deleted from a table the table manager attempts to balance the indexes for the table However an optimum balance cannot always be maintained Running tqrorg causes the indexes to be rebuilt ensuring that they are properly balanced Properly balanced indexes increase the efficiency of record searching In the event of an abnormal system termination it is possible that one or more table files will be left in an inconsistent state Running tqrorg can repair these problems It is possible some of the most recently added records can be lost during this recovery process 11 27 Customizing Tables 11
288. ge is created by tqalm if a threshold violation has occurred on your system for fixed tracks available TQ CRITICAL lt 43212 000 gt CLOI Fixed Trks Available CRITICAL indicates the alarm severity and 43212 000 is the current value for the parameter CLO1 is the name of the system on which the alarm occurred Fixed Trks Available is the alarm identifier that was specified when the threshold definition was created using tqzip In this example quall qual2 and qual3 do not exist for this parameter 7 19 Site Specific Alarm Handling You can perform site specific actions to handle alarms by writing a C language function that is called by the tqalm program For each alarm that is generated tqalm calls a function named usralarm It passes the address of a data structure to this function that contains information on the alarm that has just been generated For a description of this data structure you can look at the SYS LIB TQ BASELINE USRALARM C element By default the usralarm function returns to tqalm without performing any action on the alarm To add site specific code you must update SYS LIB TQ BASELINE USRALARM C and compile it using the UC compiler You then need to copy the USRALARM C object module created by the UC compiler to TPF and enter the following control statement to relink tqalm LINK SYS LIB TQ BASELINE LNK USRALARM SYS LIB TQ BASELINE TQALM You must have write access to SYS LIB TQ BASELINE to put the new tqalm
289. gend right Specifies the relative location for the right coordinate of the bottom right hand corner of the window containing the legend Legend Style Font Style Size Foreground Background Border The specifications for the text attributes of the entries of the legend Font Specifies the typeface in which the text is printed Style Specifies the kind of emphasis for the typeface of the text Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the text be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Defines the number of points for the typeface of the legend text This value determines the height of the characters in which the legend text is printed The value must be a positive integer Foreground Indicates the color in which the text of the legend is printed C 20 TQ 01011 18 Report Definition File Background Indicates the color of the area that appears directly behind the legend The background area is a boxed area and acts as a frame for the legend If a background color is specified the boxed area behind the legend covers any graph lines and other text that may appear behind the legend If the Background field for the legend text is not specified TeamQuest View al
290. gregation 6 27 recovery 10 8 removing aggregation set 6 19 saving derived statistic 8 13 dedicated Java processor statistics A 14 default threshold definitions 3 4 changing 3 5 using 3 4 defining named pipe connection 4 6 retention specifications for tables 11 9 socket connection 4 5 socket session 4 1 SSL connection 4 6 definition rules for workloads 5 2 deleting aggregation set 6 15 6 19 an alias 8 13 collection period for tables 11 8 TQ 01011 18 Index derived statistic 8 17 records from tables 11 2 11 16 retention specifications for tables 11 14 threshold definitions 7 19 workload definition 5 18 workload set definition 5 12 deregistering tables 11 20 derived statistics alias definition 8 2 creating 8 8 creating thresholds for 7 14 defining an alias 8 11 definition file format 8 18 deleting 8 17 description 2 6 displaying 8 16 exporting definitions 8 15 exporting definitions toa file 8 18 formula definition 8 2 8 3 hierarchy 8 6 identification of 8 1 importing definitions 8 15 loading definitions 8 21 maintenance 8 5 modifying 8 16 parameter references 8 2 parameter substitution 8 2 saving 8 13 TeamQuest Performance Software Administration 8 4 tqzip administration 8 17 unloading definitions 8 18 updating thresholds for 7 16 using tqrpt for reporting D 5 viewing thresholds for 7 18 workload performance A 51 disk space requirements 10 4 displaying
291. gregation set TQ 01011 18 Customizing Aggregation Sets e Date consolidation The second dimension of consolidation is date consolidation For example you might have data that is collected at 30 minute intervals throughout each day A single data point exists for each 30 minute interval You can create a second aggregation set that still has a granularity of 30 minutes but that has a group type of a week In this second aggregation a single data point represents the 30 minute interval for the week In the first aggregation set the day consolidation type the data point for 10 00 to 10 29 represents the value for that interval for the day In the second aggregation set the week consolidation type the data point for 10 00 to 10 29 represents the value of the seven data points for each day of the week When you are consolidating data by days you might want to ignore certain days of the week so that you are accurately representing the normal work days of the system users For this reason TeamQuest Baseline allows you to exclude certain days of the week when you consolidate data from a day consolidation type into a week or month consolidation type When you create such an aggregation set you must specify the days to ignore in the aggregation set definition TeamQuest Baseline normally assumes that all days will be included unless you specify the days to exclude TeamQuest Baseline allows you to specify the day of the week that should
292. haracteristics Prg_Num Appl_Group ISPEC_Name PID Txn_Code Txn_Run_ID Logical lt Operators Operators 2 Select the characteristic and click the arrow button The characteristic then appears in the Workload Characterization edit box 3 Select the operator from the Operators list box and click the arrow button The operator appears in the Workload Characterization edit box 4 Enter the numerical condition or string condition to complete the definition If you want multiple definitions use the Logical Operators list box to select an operator and enter another workload definition statement You can also double click the characteristics operators and logical operators to move them between the list boxes and the Workload Characterization edit box 5 14 TQ 01011 18 Customizing Workloads Workload Definition Dialog Box The Workload Definition dialog box contains information about the session database source data type and workload set name for which you are creating a workload definition It contains the following fields Workload Name supply a name for the workload The workload name is limited to 51 characters in length Workload Characterization enter the workload definition The workload definition must be expressed using the workload definition language You can type in the definition directly into the edit box or you can use the Characteristics Operators and Logical Operators list boxes See 5 2 2 for
293. hasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the title be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Defines the number of points for the typeface of the page title This value determines the height of the characters in which the page title is printed The value must be a positive integer TQ 01011 18 c 9 Report Definition File C 10 Foreground Indicates the color in which the page title of the report is printed Background Indicates the color of the area that appears directly behind the page title of the report The background area is a boxed area and acts as a frame for the page title of the report If a background color is specified the boxed area behind the page title covers any graph lines and other text that may appear behind the page title If the Background field for the page title is not specified TeamQuest View allows any graph lines or other text that may be behind the page title to show through Border Indicates the color of the border line that appears for the background area of the page title If the Border field is not specified TeamQuest View does not display a border line around the background area of the page title The default values for Title Style are defined in the Report Styles settings of the User Preferences See the TeamQuest View User Guid
294. hat has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 g specifies the granularity for the aggregation set The granularity parameter is expressed as n nsec nmin or nhour where n is the number of seconds minutes or hours Do not enter a space between the number and the unit specification If no unit is specified seconds is assumed The default is 30 minutes S specifies the start time of the aggregation set The starttime parameter is in the form of hhmm ss The default is 000000 e specifies the end time of the aggregation set The endtime parameter is in the form of hhmm ss The default is 235959 The time interval specified by the starttime and endtime parameters cannot exceed 24 hours However you can specify an aggregation set that starts at 070000 and ends at 305959 thus reflecting a time interval beginning at 7 00 00 a m and ending at 6 59 59 a m the next day t specifies the group type of the aggregation set Valid values for the grouptype parameter are day week and month If no group type is specified day is assumed TQ 01011 18 TQ 01011 18 Customizing Aggregation Sets specifies the amount of data that is kept for the aggregation set The keepcount parameter is expressed as n nmin or nhour where n is the number of minutes hours or groups for which the data is stored in the database Do not enter a space between the number
295. hat is selected from the list All operations are performed relative to a selected key name in the Derived Statistics list box The Open button opens a selected statistic key to display or modify the derived statistic definition The New button allows you to create a new derived statistic definition The Delete button allows you to delete a derived statistic from the Derived Statistics list The Expand button allows you to expand the statistic hierarchy to the next key level The View button allows you to look at long key names that may have been truncated in a list box You must select a statistic key before performing any operations with the Open and Delete buttons When you click OK your changes are made to derived statistic definitions including any changes made to derived statistics attributes accessed through the New or Open buttons see 8 3 3 and 8 3 8 When you click Cancel any changes you have made are ignored Derived Statistic Hierarchy Statistics are stored in the database in a hierarchical format A statistic is defined by a hierarchy of key names which identify the statistic for retrieval from the database See 8 2 1 for a description of this hierarchy Each layer in the hierarchy is maintained by a key name The first layer is the first key the next layer is the next key and so on Derived statistic maintenance provides four layers of keys the category group the category the subcategory and the statistic All four keys mu
296. he aggregation sets Aggregation sets contain the data points that are to be saved for the period of time specified by the aggregation set definition The aggregation set files have a naming convention of TQDATA xxxxaaaaaaaa where xxxx is the first four characters of the database name and aaaaaaaa is the first eight characters of the aggregation set name Every time an aggregation set is created an aggregation set file is also created When you view an aggregation set definition with TeamQuest Performance Software Administration the File Size field is displayed This field can be used to estimate your sizing requirements It contains the approximate size in tracks that the aggregation set can attain if no new parameters are added to the database The File Size field is not shown until a probe has put at least one point of data into the aggregation set This number can change if the probe creates additional statistics or if additional probes are used See 6 3 3 for more information on viewing aggregation set definitions You can significantly reduce the disk space required to store aggregation set data by using compact aggregation sets See 6 2 5 for a description of compact versus expanded aggregation sets Listing Database Parameters You can use tqzip with the o option to export a list of database parameters to a file Command Line Format To generate a list of database parameters using tqzip use the following command line format tq
297. he database i specifies the location of the input parameter table name to be used by the tqsys probe For the tqsys probe the parameter table is the element SYS LIB TQ BASELINE SYSINPUTTBL This is a required parameter on the tqsys call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name a specifies the aggregation set to be used for data collection The aggset1 parameter is the name of an aggregation set defined in the database name The aggset1 parameter is case sensitive The aggregation set is optional on the call line If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used See 2 2 1 for a description of aggregation sets Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent lines of the processor call The parameters specified on the processor call must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keywo
298. he default values for XAxis Labeling are defined in the Axis Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Page Style Font Style Size The specifications for the text attributes of the text on the page bar of the report window Font Style Size Specifies the typeface in which the page bar text is printed Specifies the kind of emphasis for the typeface of the page bar text Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the page bar text However the styles must be joined by a plus sign For example to specify that the page bar text be printed in bold and italic typeface you enter bold italic in the Style field of the specification Defines the number of points for the typeface of the page bar text This value determines the height of the characters in which the page bar text is printed The value must be a positive integer Status Style Font Style Size The specifications for the text attributes of the text on the report status bar of the report window Font Style Specifies the typeface in which the report status bar text is printed Specifies the kind of emphasis for the typeface of the report status bar text Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can sp
299. he default values for these parameters as provided in the tqxmr runstream To modify any of the defaults you need to update the tqxmr runstream in SYS LIB RUN TQ 01011 18 9 29 Customizing Data Collection Probes 9 30 The tqmpr processor call has the following format tqmpr z L c i parameter table d database name L a aggsetl p Mapper name END Z is used to register the tqmpr probe in the database C allows tqmpr to process MAPPER accounting files that contain Biller 1100 formatted entries i specifies the location of the input parameter table name to be used by the tqmpr probe For the tqmpr probe the parameter table is the element SYS LIB TQ BASELINE MPRINPUTTBL This is a required parameter on the tqmpr call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name a specifies the aggregation set to be used for data collection The aggset1 parameter is the name of an aggregation set defined in the database name The aggset1 parameter is case sensitive The aggregation set is optional on the call line If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used See 2 2 1 for a des
300. he following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A To update a threshold enter u at the prompt B The tqzip program displays a list of the parameters and derived statistics that currently have thresholds defined Select the derived statistic threshold definition that you want to update You can enter either the number or the full name of your selection Derived statistic names begin with DERIVED C The derived statistic qualifier lists are displayed and you are given the chance to update your selections To add a selection to the qualifier list you enter the number or name of the qualifier value To remove a selection from the list you enter the number of the qualifier values preceded by a hyphen An asterisk is used to indicate which qualifier values in the list have been selected You continue adding or removing selections in the qualifier list until entering an E to exit D Note the usage of the hyphen to remove IP001 from the qualifier list E After you select a parameter name the tqzip program displays the series of threshold definition prompts with the current value set as the default for each prompt see G through P in 7 4 You can modify any of the fields TQ 01011 18 7 17 Customizing Alarms and Thresholds 7 10 Viewing Derived Statistic Threshold 7 18 Definitions The tqzip program allows you to view an existi
301. he operating system Q Quota The system resources that a run is allowed to use The Exec can be configured to maintain Quota limits for each account number and user id Quota set A method of resource control used by the Exec Quota sets limit the resources used by the attached account or user id You can attach Quota sets to accounts and user ids The Exec enforces the limits defined by the Quota set if the Quota level in the Exec is set SIP 1 See Software Instrumentation Package 2 The Quota privilege to allow all SIP functions This privilege is required for the account running TeamQuest Baseline or TeamQuest Online on the host SIPONOFF The Quota privilege to allow turning SIP on and off by using the ER SYSBAL This privilege is required for the account running TeamQuest Baseline or TeamQuest Online on the host SIPREDCT The Quota privilege to allow SIP data reduction This privilege is required for the account running TeamQuest Baseline or TeamQuest Online on the host Software Instrumentation Package SIP An Exec system performance data collection function SIP is configured during a system generation and then activated or deactivated by the operator at the system console by using system balance keyins Software Library Administrator SOLAR An OS 2200 software product used to load OS 2200 software products from a software package tape You must use SOLAR to install TeamQuest Baseline and TeamQuest Online Glossary
302. he socket names that have already been used with a session Each time you define a new session the host name or IP address is added to this list For more information see 4 4 2 4 4 TQ 01011 18 TeamQuest Performance Software Administration Named Pipe Settings These fields are used to specify the settings for named pipe connections Host Name or IP Address This field contains the name of the host or the IP address of the system where the performance database resides Named Pipe list This list box contains a list of the names of the named pipes that have already been used with a session Each time you define a new session the host name or IP address is added to this list For more information see 4 4 3 4 4 1 Using a Socket Connection When you select the Socket option button from the New Session dialog box you need to specify a host name or an Internet Protocol IP address and a port number You can either enter a host name or IP address in the edit box or you can choose from the current list of hosts below the edit box that are already defined as part of a session You may need to specify the full domain name of the host to which you want to connect Each time you define a new session the host name or IP address is added to the list of currently defined hosts Note Before using sockets you must make sure your host software is configured to accept socket connections on that host See Appendix I for more information on s
303. he value title to show through Border Indicates the color of the border line that appears for the background area of the value title If the Border field is not specified TeamQuest View does not display a border line around the background area of the value title The default values for Value Style are defined in Report Styles settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Value Axis Enable Font Style Size Color The specifications for the text attributes of the labels for the vertical axis The axis labels appear along the axis at the major tick marks and are used to identify the values of the vertical axis TQ 01011 18 Enable Font Style Size Indicates whether the labels appear along the vertical axis The acceptable values are ON or OFF If Enable is set to ON the labels appear on the report If Enable is set to OFF the labels do not appear on the report The default for this field is ON Specifies the typeface in which the labels are printed Specifies the kind of emphasis for the typeface of the labels Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the labels However the styles must be joined by a plus sign For example to specify that the labels be printed in bold and italic typeface you enter bold italic in the Style field of the specific
304. hold definition for the alarm A count of alarms that is used for consolidating the alarms at report time The name of the parameter for which the alarm was generated This includes system category group category subcategory statistic resource application group workload set and workload The name of the performance database that contains the data that was used to detect this alarm condition TQ 01011 18 Appendix H tqzip Processor Call Format The tqzip program has the following processor call format tqzip where h a 0 P S u TQ 01011 18 h L d databasename a action o parm file p parm file 1 u L e c tablespec file r reduction file s derived file t threshold file w workload file prints usage information specifies the database that tqzip should use The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the default database is used indicates the action to be taken with the threshold checking setting action can be one of the following ON turns threshold checking on OFF turns threshold checking off LIST displays the current threshold checking setting See 7 15 for more information on turning on and off threshold checking outputs the database parameter list to the file specified by the parm file parameter See 10 8 deletes parameters from the
305. hoose from the current list of hosts below the edit box which are already defined as part of a session You may need to specify the full domain name of the host to which you want to connect Each time you define a new session the host name or IP address is added to the list of currently defined hosts Note Named Pipe connections are supported for connecting to Unisys systems only If you need to remove host names or IP addresses select the item you wish to remove from the list and press your keyboard Delete key Additional Sessions If you have multiple hosts collecting data you can define and open additional host communications sessions to maintain your workloads aggregations and derived statistics To open additional sessions select File gt New Session and follow the instructions for creating a new session see 4 4 TQ 01011 18 TeamQuest Performance Software Administration 4 6 Session Timeout Warnings Warning screens may appear during your administration session When a request is made to the host and the host does not acknowledge the request within the specified timeout period TeamQuest Performance Software Administration displays the Session Timeout window The following two messages can be displayed in the Session Timeout window Session has not acknowledged request Host database server has still not acknowledged a data request sent to it within the sessions communications timeout time Session sent an ackno
306. ial identifiers amp BTIME The beginning time of the report The format of the time is the locale default format for time amp BDATE The beginning date for the report The format of the date is the locale default format for date amp BTDATE The beginning time and date for the report The format of the time and date is the locale default format for time and date TQ 01011 18 Host Based Reporting for Table Data amp ETIME The ending time for the report The format of the time is the locale default format for time amp EDATE The ending date for the report The format of the date is the locale default format for date amp ETDATE The ending time and date for the report The format of the time and date is the locale default format for time and date amp NTIME The current now time The format is HHMMSS which is suitable for report time range specification amp NDATE The current now date The format is YYYYMMDD which is suitable for the report date range specification amp NTDATE The current now time and date The format is YYYYMMDD HHMMSS which is suitable for the report time range specification amp PAGE The current page being generated amp USER The login name of the user generating the report amp SYSTEM The name of the system where the report is being generated amp TABLE The name of the table for which data is being printed TQ 01011 18 E 11 Host Based Reporting for Table Data
307. ialog box see 6 3 1 and click Open You can also double click on the aggregation set name in the Aggregation Sets list box The Aggregation Set Attributes dialog box is displayed Aggregation Set Attributes Session Database Agg Set Name Granularity Source Agg Set Retain Data For Date Consolidation Type First Day cl01 PRODUCTION 5 minute 00 05 00 1 minute 1 Week Week z Monday v Exclude Expanded Agg Set Data Collection Times Range 07 00 00 14 59 59 mj mi Interval Make your changes as required For a description of the fields in this dialog box see 6 3 2 When you click OK your changes are accepted and you return to the Aggregation Set Maintenance dialog box see 6 3 1 Your changes are not sent to the host software until you click OK in the Aggregation Set Maintenance dialog box see 6 3 1 6 14 TQ 01011 18 6 3 4 6 4 Customizing Aggregation Sets Deleting an Aggregation Set To delete an aggregation set from a database select an aggregation set name from the Aggregation Sets list box in the Aggregation Set Maintenance dialog box see 6 3 1 When you click the Delete button an alert box verifies your request to delete the aggregation set When you click Yes in this alert box the aggregation set name is removed from the Aggregation Sets list box in the Aggregation Set Maintenance dialog box see 6 3 1 To delete the aggregation set definition from the
308. ic Name Hardware Default Unretryable Hwr Chk Diagnostic Interrupt 3 Interrupt 4 Interrupt 5 Interrupt 6 Interrupt 7 Reference Violation Addressing Exception Terminal Addr Exception RSC Underflow Overflow Signal Test and Set Invalid Instruction Page Exception Arithmetic Exception Data Exception Operation Trap Breakpoint Quantum Timer Interrupt 21 Interrupt 22 Interrupt 23 Software Break Jump History Full TQ 01011 18 Percentage of TIP processing Minimum percentage of TIP processing Maximum percentage of TIP processing Percentage of critical deadline processing Percentage of demand processing Percentage of batch processing Processor Java Total Hardware Interrupts Hardware default interrupts per second Unretryable hardware check interrupts per second Diagnostic interrupts per second Interrupt 3 not used per second Interrupt 4 not used per second Interrupt 5 not used per second Interrupt 6 not used per second Interrupt 7 not used per second Reference violation interrupts per second Addressing interrupts Terminal addressing exception interrupts per second RCS generic stack underflow overflow interrupts per second Signal interrupts per second Test and set interrupts per second Invalid instruction interrupts per second Page exception interrupts per second Arithmetic exception interrupts per second Data exception interrupts per second Operation trap interrupts per second Breakpoi
309. ield values that is the values following the colons as long as you keep the fields in their same relative position on each line For example if you are modifying the descriptive information of line 3 do not alter the text Number of zones or Normal zone You can however change the field value for the number of zones from 3 to 2 The spaces surrounding field values are insignificant You can add additional threshold definitions by duplicating these lines and modifying the parameter name line and the descriptive information for the new threshold The last line of the thresholds file must be the END THRESHOLDS line You must not change or delete this line 7 23 Customizing Alarms and Thresholds 7 14 Loading Threshold Definitions 7 24 You can reload the performance database with a new or updated set of threshold definitions by using the l and t options on the tqzip processor call You should use a file as created by the u and t options see 7 13 as input for loading threshold definitions If you update the threshold definition file you should ensure that the format of the file remains consistent with what was output by the u and t options Notes e The tqzip program deletes any existing threshold definitions from your database before loading the new definitions Use the e option on tqzip if you do not want existing definitions deleted e After the threshold definitions have been loaded into the database all
310. ields should be the first fields in the table In the example Time Interval and System are the fields that are part of the group key You should use these field definitions in most cases They should be the first three fields defined A unique group key normally defines one sample interval of data 9 50 TQ 01011 18 Customizing Data Collection Probes e Primary Key Fields Because you are storing table oriented data each sample interval normally has more than one record with each record describing a unique entity that is being measured Therefore each table must have one or more primary keys that uniquely define the entity being measured or reported Group keys are almost always primary keys as well However you usually need one or more additional fields as primary keys to uniquely define each record in the table Normally this will be a single field that describes the entity being measured or reported such as User or Account e Data Fields Data fields are the fields that normally contain the measurement data about the entity being monitored Field Definition There should be one field definition for each field column in the table Each field definition must be on a single input line and contain the following information FIELD Field_Name field name Group_Key Primary_Key Identifier field type data type data subtype Max_Size max size LTime_Consolidation Seq_Cons_Method
311. ies the hatch pattern that is used for a particular graphic object Valid hatch patterns are Horizontal For a horizontal line pattern within the object Vertical For a vertical line pattern within the object ForwardDiagonal For a line pattern of stripes on a forward diagonal within the object BackwardDiagonal For a line pattern of stripes on a backward diagonal within the object Cross For a line pattern of square shapes within the object DiagonalCross For a line pattern of diamond shapes within the object C 27 Report Definition File Marker Specifies the marker type for a point or line graph Valid markers are Invisible For no marker Diamond For a diamond shaped marker Triangle For a triangle shaped marker Square For a square shaped marker Circle For a circle shaped marker Line Specifies the line type on the line graph The valid line types are solid Dot Dash Medium or Heavy C 28 TQ 01011 18 Report Definition File C 5 Parameter Group Sections A parameter group section appears once for each parameter group that is displayed in the report The parameter group section explains database information used in the data request to the host Each parameter group represents an individual data request A parameter group can create multiple parameter lines Format The format for the parameter group section is parameter group Keyname keyvalue keyvalue Value Types Average StdDev All Top
312. ified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name a specifies the aggregation set to be used for data collection The aggset1 parameter is the name of an aggregation set defined in the database name The aggset1 parameter is case sensitive The aggregation set is optional on the processor call If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used See 2 2 1 for a description of aggregation sets 9 26 TQ 01011 18 Note Customizing Data Collection Probes specifies an IOTRACE log file to be processed JOTRACE log is the name of the log file n is the file cycle number When processing multiple IOTRACE log files multiple f options are needed one for each cycle to be processed All cycles must be in sequential order and from the same IOTRACE session You can process IOTRACE tape files that have been previously assigned to your run by specifying the file name used to assign the tape specifies the system name that identifies the system source of the IOTRACE log file data It is used to determine where to store the data in the performance database The system name parameter is case sensitive Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent l
313. ight process the previous day s data and populate the 10 minute aggregation set that is aggregated when tqagg runs Both probes process TPM workload data based on the same workload sets For the tqrtpm probe the workload sets can optionally be specified on the call line to limit the amount of workload processing required and reduce processing overhead Depending on the volume of transactions that occur the tqrtpm probe may fall behind while reporting data in near real time For this reason it is recommended that the tqrtpm probe be run with the r option to enable it to run at a real time priority Both probes require that your TPM audit trail be configured to mass storage or moved to mass storage prior to starting These probes report on information from the TPM TIP termination records type 13 subtype 3 generated when TIP is configured in the Exec By default audit trail 17 is referenced for TPM log file selection For example the ACI SYS ACI 17 and SYS F017L1 file cycles The audit trail number can be specified on the tqrtpm or tqtpm processor call statement By default the tqtpm program processes all cycles of the TPM log files from the previous day During the tqinit process at installation time you are prompted for the time that the tqtpm runstream starts each day The tqtpm probe should run each day to ensure that your performance database is kept up to date Both probes also analyze a common format of the user area common to so
314. igits and the plus sign if it precedes the numeric digits Any other character will cause the field and thus the record to be rejected The value must be a valid unsigned long integer Subtype is not checked for this type Allowable characters are numeric digits the plus sign and hyphen if they precede the numeric characters or follow the exponent character the period and the e and E characters if they follow the numeric characters and have an optionally signed value following Any other character will cause the field and thus the record to be rejected The value must be a valid floating point value Subtype is not checked for this type Allowable characters are numeric digits the plus sign and hyphen if they precede the numeric characters or follow the exponent character the period and the e and E characters if they follow the numeric characters and have an optionally signed value following Any other character will cause the field and thus the record to be rejected The value must be a valid double precision floating point value Any valid characters If the string contains spaces it must be enclosed in double quotation marks Any valid characters If the string contains spaces it must be enclosed in double quotation marks 9 67 Section 10 Customizing Performance Databases 10 1 Introduction The tqdbu and tqzip programs are used to administer your performance databa
315. ility in tqzip to assist in copying threshold definitions from one system to another see 7 13 and 7 14 TQ 01011 18 7 33 Section 8 Customizing Derived Statistics 8 1 8 2 8 2 1 Introduction TeamQuest Performance Software Administration or the tqzip program can be used to administer the derived statistics in your performance database If you want to customize your derived statistics you can perform the following tasks e Create derived statistics e Delete derived statistics e Modify derived statistics e View derived statistics Concepts Some basic concepts are essential to the understanding of TeamQuest Baseline derived statistics You should have an understanding of these concepts before beginning your administrative tasks with derived statistics Derived Statistic Identification A derived statistic is identified in the database using the TeamQuest Baseline key name hierarchy as described in 2 2 2 Derived statistic definitions apply to all systems in your database Because of this a derived statistic is actually identified by using only the category group category subcategory and statistic key names When you create a derived statistic with TeamQuest View you are allowed to insert it at the statistic level at any point in the existing parameter hierarchy A special location exists in the hierarchy if you want to group all of your derived statistics together This special location is identified by Category
316. imary database You can then delete your test databases using tqdbu 9 11 1 Writing a User Probe for Performance Data The tqusrprb probe uses the information from two user created files and the parameters on its processor call to store user statistics into the performance database The two files consist of a control file and a file that contains the data to be stored in the performance database Parameters on the tqusrprb processor call line determine where the data is stored in the selected database There are two methods of calling the tqusrprb probe calling tqusrprb probe each time you want to store a data sample or running the tqusrprb probe continuously processing multiple file cycles of PRBDATA as they become available The characteristics and setup of the TQARG and PRBDATA files apply to both methods of calling the tqusrprb probe and are described below See Section 9 11 1 2 for information on running the tqusrprb probe in continuous mode TQ 01011 18 9 39 Customizing Data Collection Probes To write your own probe for performance data you need to create the control file TQARG and write an addstream or a program in the programming language of your choice to collect your statistic values and to create the data file PRBDATA The addstream or program must generate ASCII plain text output that matches the format of the input as expected by the tqusrprb probe The tqusrprb probe reads the statistic names and aggregation weights as
317. imun line width in characters In comma delimited output each of the fields are separated by commas In tab delimited output the fields are separated by tab characters A maximum of 255 fields are allowed on each output line Since date and time take up two fields you can have up to 253 parameters in a report using the x option The w option of tqrpt also controls the maximum width of an output line For condensed output the default line width is 4 000 characters By using the w option the line width can be decreased or it can be increased to a maximum of 8 000 characters When either the field or character maximum is reached on a line any remaining output for that line is discarded Truncation of a line will not result in partial field values For example if the last field of a line will not entirely fit within the line limits the complete field is truncated for all lines on the report The condensed report output consists of a set of a parameter definition lines followed by a series of detail lines The parameter definition lines contain the names of the parameters selected for the report The parameters are listed in the order that they were selected with each parameter key name on a separate line The detail lines contain the data values for each of the parameters for the selected time period Each detail line begins with date and time fields that identify the data values on that line For example the following parameters were selected
318. in the preceding example The letters are not part of the actual screen display A The tqzip program prompts you for the number of the retention that you want to delete If you do not know the number of the retention enter a question mark to list the currently defined retentions for the table B The tqzip program displays the list of retentions for the table c The tqzip program prompts you again for the number of the retention you want to delete In this example retention number 1 1 minute data retained for 1 week is deleted After the deletion data will be retained at the original collection interval for 2 days then aggregated into 15 minute records that will be retained for 1 week 11 14 TQ 01011 18 Customizing Tables 11 6 5 Setting the Retention Start Time The retention start time is the synchronization time for aggregated records All aggregated records will start on even intervals of the retention start time In most cases the retention start time can be left at 00 00 However the tqzip program allows you to change the retention start time if needed To set the retention start time select Set Retention Start Time from the Retention Specification Menu see 11 6 by entering 5 at the prompt The tqzip program then displays the following prompt gt Enter your selection lt 1 gt gt 5 gt gt Retention Start hh mm lt 00 00 gt gt 8 00 Enter the new start time in hh mm format The retention st
319. ines of the processor call The parameters specified on the processor call must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keyword The tqiot program processes each IOTRACE log file storing data points in the performance database as specified by the aggregation set parameter a and the system parameter s TQ 01011 18 9 27 Customizing Data Collection Probes 9 8 3 9 28 Workload Characteristics for the IOTRACE Probe The IOTRACE probe has a unique set of characteristics that are used for its workload definitions You use these characteristics when you create the workload definitions using TeamQuest Performance Software Administration see 5 3 The following characteristics define the IOTRACE workloads Characteristic Valid Type and Valid Name Characters Max Length Operations Run_ID A Z 0 9 string 6 l Device A Z 0 9 string 6 File A Z 0 9 string 12 TIP_File 0 9 string 6 i Program A Z 0 9 string 3 RS Transaction A Z 0 9 string 12 l Example 1 The following example shows a workload definition using run identifiers to identify system jobs This workload contains information for system background runs MSCP RUNNER CMR PCCTI and DDP By using the Run_ID MSCP Run_ID MSCP construct data for runs with identifiers of MSCP and MSCPx
320. ing SMART SP AMS e Alarm Identifier Each threshold should have a unique alarm identifier with no embedded spaces The unique alarm identifier reduces the possibility of an unintended pattern match By default SMART SP AMS breaks messages into tokens using spaces as delimiters By ensuring that the alarm identifier does not contain spaces it makes extracting the identifier token easy and adds some measure of predictability to the message e User Text Lines User defined alarm text can be defined as three separate messages using tqzip Each of these messages is displayed on a separate line in the TeamQuest Baseline alarm message This causes multiple output lines on the console Since SMART SP AMS requires you to use variables to correlate multiple line console messages you can simplify the pattern definition and avoid the use of variables by defining any user text for the alarm within SPO itself If you define multiline console messages with tqzip then putting the unique alarm identifier on each line along with a sequence number will reduce the possibility of an unintended pattern match The user text lines within TeamQuest Baseline can still be used if you want the user text to appear in the alarm message generated by TeamQuest Baseline It is important to remember however that because this user text is not on the first line of the alarm message it will not be available within the SMART SP AMS pattern matching done for the alarm T
321. inition stored in a file or by using one of the default reports definitions provided for the table The tqrep program is run on the host system and allows you to generate reports without using the TeamQuest View user interface Figure E 1 shows an example of the type of report that can be generated using tqrep a Alarm Summary for 01 04 12 N Timestamp Severity Alarm Id Value Qual 1 Qual 2 Qual 3 0104 10 14 CRITICAL usrcpu 2 53 0104 10 21 CRITICAL syscpu 15 28 0104 10 23 NORMAL syscpu 1 75 0104 10 24 CRITICAL syscpu 5 42 0104 10 25 NORMAL syscpu 4 38 0104 10 26 CRITICAL syscpu 1523 0104 10 30 NORMAL syscpu 4 92 0104 10 51 CRITICAL syscpu 7 65 0104 10 57 NORMAL syscpu 1 33 Figure E 1 Report Generated by tqrep TQ 01011 18 E 1 Host Based Reporting for Table Data E2 E 2 Command Format The tqrep command line has the following format tqrep h L d databasename 1 s starttime e endtime c time consolidation t tablename r z n defRepname f reportdefnfile o outputfile x delimiter L w reportwidth END where h d l S prints usage information containing the command line syntax specifies the database from which you want to generate a report The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the default database is used
322. ion specifications In this example there are four specifications The column headings for the retention specifications indicate the number of the specification the interval at which the data is collected and the length of time the data is kept in the database 8 14 11 Customizing Tables D The list of retention specifications shows the current retentions The retentions are listed in the order in which the data is retained The primary retention specification is always listed first followed by the additional retention specifications In this example there are four retentions The data in the primary retention is kept for 1 day at whatever interval it is collected Retention number 1 stores data at 1 minute intervals for 1 week Retention number 2 stores data at 30 minute intervals for 1 month Retention number 3 stores data at 1 hour intervals for 6 months In terms of storing the data the data in the primary retention is kept for one day At the end of the day the aggregator aggregates the original records into 1 minute intervals and removes the records that were aggregated The 1 minute interval data is kept for 1 week At the end of a week the aggregator aggregates the 1 minute records into 30 minute intervals and removes the 1 minute interval records that were aggregated The 30 minute interval data is kept for 1 month At the end of a month the aggregator aggregates the 30 minute records into 1 hour intervals and removes
323. ip command line options you can perform the following tasks e Unload derived statistic definitions contained in the database to a file see 8 4 1 You can use this file to view the definitions and update them by using a text editor You can then reload the definitions e Load derived statistic definitions from a specified file into the database see 8 4 2 TQ 01011 18 8 17 Customizing Derived Statistics 8 4 1 Unloading Derived Statistic Definitions The tqzip program allows you to export all the derived statistic definitions currently contained in the performance database into a file You can use this file to view the derived statistic definitions and update them by using a text editor You can then reload the derived statistic definitions by using the l and s options on the tqzip processor call see 8 4 2 Note It is recommended that you use the Administration utility to create and update your derived statistic definitions Using the tqzip processor call options to unload and load a derived statistics file provides an alternative way to update definitions To unload derived statistic definitions with tqzip use the following processor call format tqzip d databasename u s derived file where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used se
324. irements See Section 6 for more information on maintaining the aggregation sets of your database 2 Configure workload definitions Workload definitions may be a consideration at your site Workloads provide the capability to group user system utilization This is meaningful in determining usage by each group for example development 20 administration 15 database 58 and other 7 Workloads provide a method of determining the resource utilization for a given group of users See Section 5 for information on maintaining the workloads for TeamQuest Baseline 3 Configure alarms and thresholds to send alerts to the console based on user defined thresholds for system parameters This is very helpful in monitoring systems that require minimal service as well as systems operating with restrictive limits Alarms may send console messages and call a user supplied routine for site specific alarm handling See Section 7 for more information about defining alarms and thresholds for your system 4 Create derived statistics Derived statistics are user defined formulas made up of collected parameters For example the parameters collected from the system may not provide an overall average statistic and by using a combination of collected statistics you can define a derived statistic to represent the average Derived Statistics are stored as formulas in the database and are calculated at report time Derived statistics can be added to any report or inclu
325. is also allowed For example 2 0e 3 is the same as 002 Arguments for mathematical functions are specified as a list of values separated by commas and enclosed in parentheses Each value can be a numeric constant or a parameter reference alias You cannot nest mathematical functions by specifying a function name in an argument list See 8 2 3 for the mathematical functions supported by TeamQuest Baseline 8 10 TQ 01011 18 Customizing Derived Statistics 8 3 3 1 Defining an Alias To define an alias press the Define button next to the Alias List edit box in the Derived Statistics Definition dialog box The Define Alias dialog box is displayed Define Alias m Statistic Hierarchy PRODUCTION Derived Processor 10 Activity Memory System Log H STAT Log Mapper Log System UDS MassStor by System MassStor by Equip MassStor by Disk Define Alias Dialog Box The Define Alias dialog box contains the following fields Alias provides a short name for the alias This field is limited to 12 characters in length and must begin with an alpha character from the set A through Z and a through z The rest of the characters must be alphanumeric Statistic Hierarchy contains a list of all the statistics defined in the database you are working on including derived statistics The statistics are listed in the hierarchy described in 8 3 2 A status box under the list box indicates the key name type category group category sub
326. is value determines the height of the characters in which the labels are printed The value must be a positive integer Color Specifies the color in which the horizontal axis and the labels are printed The default value is Black The default values for XAxis are defined in the Axis Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences XAxis Labeling Major Minor Midnight The specifications for the labels of the horizontal axis Major Determines if labels appear at the major tick along the horizontal axis The acceptable values are ON or OFF If Major is set to ON TeamQuest View displays labels at the major ticks If Major is set to OFF TeamQuest View does not display labels at the major ticks Minor Determines if labels appear at the minor ticks along the horizontal axis The acceptable values are ON or OFF If Minor is set to ON TeamQuest View displays labels at the minor ticks If Minor is set to OFF TeamQuest View does not display labels at the minor ticks Midnight Determines if a label appears when the time interval crosses midnight Acceptable values are ON or OFF If Midnight is set to ON a date label appears when the data crosses midnight If Midnight is set to OFF the date label does not appear when the data crosses midnight This field is used when the date and time points are graphed on the x axis TQ 01011 18 C 17 Report Definition File C 18 T
327. ist of the parameters collected by the TeamQuest Baseline tqmips probe for Capacity On Demand COD systems Note These parameters are not collected by TeamQuest Online Parameter Hierarchy Category Group Category Subcategory Statistic Name Avg MIPS Used MIPS Seconds MIPS Interval Desired MIPS Baseline MIPS Ceiling MIPS Running MIPS MIPS Total Utilization Average number of MIPS used Total number of MIPS seconds used Total number of second for the reporting period The number of desired MIPS for the partition The number of baseline MIPS for the system The number of ceiling MIPS for the system Total number of running MIPS for the partition A 4 1 MIPS Probe Derived Statistics The following statistics are collected by the MIPS tqmips probe Parameter Hierarchy Category Group Category Subcategory Statistic Name Desired Percent Baseline Percent Ceiling Percent Running Percent MIPS Months TQ 01011 18 MIPS Total Utilization Percentage of the Desired MIPS used Percentage of the Baseline MIPS used Percentage of the Ceiling MIPS used Percentage of the Running MIPS used Number of MIPS Months used A 43 OS 2200 Parameters A 5 A 44 IOTRACE Parameters This subsection contains a list of the parameters collected by the TeamQuest Baseline IOTRACE probe The statistics are collected for each workload defined on the system Note These parameters are not collected by TeamQuest Online
328. ist to hold the data The table creation phase defines the characteristics of the table and the data to be stored in it You must define the characteristics of the table Table Definition and the characteristics of the data to be stored in the table Field Definitions These two types of definitions comprise the Table Input Specifications The tqtblprb reads the Table Input Specifications and creates the table Table Input Specifications A table input specification defines a new table Each table input specification consists of two parts one input line that contains the table definition and one input line for each field in the table containing that field s definition e A table definition describes the overall table characteristics and begins with the keyword TABLE e Field definitions describe the characteristics of every field in the table and begin with the keyword FIELD Table Definition The table definition must be the first line in the table specification and must contain the following information TABLE Table_Name table class table name File_Name fi e name time type Retention ndays where TABLE Defines a table input line and must be first on the line This item is required Table_Name table class table name Specifies the table name and is a required item Table_Name is the keyword for the item table class specifies the class for the table For example 2200 and Alarms can be classes of tables table n
329. istributed Transaction Processing data 9 14 of SIP data 9 2 of System Activity Monitor data 9 5 of system log data 9 33 of TPM log data 9 17 processing SIP data 1 8 starting 9 1 storing data 1 8 tqetl runstream 9 1 data display description 1 6 displaying reports 1 6 requesting data 1 6 retrieving data 1 6 data storage tqusrprb probe 9 45 data value 2 3 database aggregation set 1 10 2 1 creating 10 3 data value 2 3 deleting 10 3 description 1 10 2 1 10 1 disk space requirements 10 4 files 10 2 listing 10 4 structure 2 1 table data 2 4 database administration backup 10 8 clearing locks 10 7 creating additional databases 10 3 aggregation set 6 10 6 16 derived statistic 8 8 workload definition 5 13 workload set definition 5 10 Index 2 deleting aggregation set 6 15 databases 10 3 derived statistic 8 17 workload definition 5 18 workload set definition 5 12 displaying aggregation set 6 14 derived statistic 8 16 workload definition 5 16 workload set definition 5 11 estimating disk space 10 4 exporting derived statistic definition 8 15 workload definition 5 17 importing derived statistic definition 8 15 workload definition 5 17 listing aggregation set 6 20 databases 10 4 parameters 10 4 maintaining aggregation sets 6 8 alarms 7 1 tables 11 3 thresholds 7 1 workloads 5 8 modifying aggregation set 6 14 derived statistic 8 16 workload definition 5 16 workload set definition 5 11 performing data ag
330. it Resource Wait Active Time Cards In Runs Opened Max Real Memory Real Memory Usage TQ 01011 18 System Log Workload Batch Total SUPs used to process the job Sum of the IP SUPs ER CC SUPs operating system SUPs and I O SUPs Total IP SUPs Total ER and control card SUPs Total Exec Services I O SUPs Total Exec Services IP SUPs Total control statement I O SUPs Total control statement IP SUPs Total I O SUPs Total core block SUPs Sum of SUPs and main storage usage Total number of I O requests made by this run Total number of I O requests to disk that have been made by this run Total number of words that have been transferred during disk I O operations by this run Total number of I O requests to tape that have been made by this run Total number of words that have been transferred during tape I O operations by this run Total voluntary wait time Sum of user response time and TWAIT time Total amount of time the run spent waiting for a resource Total time the job was active Total number of cards the Exec read during the run Total number of runs that were opened Maximum number of real pages concurrently resident for this run The product of the total number of resident pages and the time in seconds that the pages are resident A 49 OS 2200 Parameters Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs ES IO SUPs ES IP SUPs CS IO SUPs CS IP SUPs
331. ite miss service time Cache disk write miss hardware service time Cache disk write miss percent of all requests IO Activity Cache Disk Read Bypass Cache disk read bypass requests per second Cache disk read bypass K words per request Cache disk read bypass transfer time Cache disk read bypass service time Cache disk read bypass hardware service time Cache disk read bypass percent of all requests IO Activity Cache Disk Write Bypass Cache disk write bypass requests per second Cache disk write bypass K words per request Cache disk write bypass transfer time Cache disk write bypass service time Cache disk write bypass hardware service time Cache disk write bypass percent of all requests IO Activity Cache Disk Store Through Cache disk store through requests per second Cache disk store through K words per request A 31 OS 2200 Parameters Transfer Time RET FLD Hardware Serv Time Percent of All Reqs Category Group Category Statistic Name Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec Deadlocks Sec Lock Reqs Unlock Reqs Delayed Resp Msg Deadlocks Category Group Category Subcategory Statistic Name Reqs Sec Transfer Time Delay Time RET Percent Busy KWords Sec KWords Req KWords Req In KWords Req Out Percent Reqs Read A 32 Cache disk store through transfer time Cache disk store through service time Cache disk store through hardware service time Cache disk store through perc
332. ive Exec call interfaces that are being utilized on the system DEM RESP RPT Demand Response Summary Lists demand response times in various time ranges DEM RUNS RPT Demand Run Summary Summarizes demand population and response information ER SUM RPT ER amp CALL Summary Provides Executive request ER and call gate counts ER TOPN RPT Most active ERs Used Shows the n most active ERs that are being utilized on the system SUMMARY RPT System Summary Gives a general overview of the system This report includes processor utilization memory utilization and I O counts TIP SUM RPT TIP Summary Shows memory and processor utilization for TIP QUE AVL RPT Available Queue Reports available queue information such as enqueued rates queue times and expansion and contraction rates TQ 01011 18 Predefined Report Descriptions B 3 6 File Name Description File Name Description File Name Description QUE SUM RPT Queue Summary Reports information on the number of queue banks number of available queues and the number of deactivations per second QUE SYS RPT System Return Queue Reports system return information including queued entries per second time spent on queues and the number of queue server activations TIP TXNS RPT TIP Txns Second Shows TIP transactions per second as well as the minimum and maximum values for the time range Summary Reports The
333. izing Aggregation Sets Expanded Agg Set allows you to specify that you want your aggregation set to be expanded instead of using the default compact aggregation set An expanded aggregation set provides more complete reporting of standard deviations than with compact aggregation sets along with increased accuracy for very large or very small data point values The advantage of compact aggregation sets is that they require much less disk space to store the data while still providing accurate data point values Date Consolidation The Date Consolidation box contains a Type list box a First Day list box and an Exclude button for specifying attributes for consolidating the data points in your data across dates Date consolidation determines if dates are consolidated into a sample interval Type list box specifies the group type for the consolidation The available options are Day Week and Month Day indicates that no consolidation of data points from other days is done Week indicates that all data points from a week for each sample interval will be combined into one point Month indicates that all data points from a month for each sample interval will be combined into one data point For example if the Type is Week the data point for the 08 00 sample interval is the average value for each day of that week at 08 00 If the Type is Month the data point for 08 00 for a specific month is the average of all values from the 08 00 sample interval fo
334. l 1 30 minute interval 2 30 minute interval 3 30 minute interval 48 Parameter Grouping for an Aggregation Set Customizing Aggregation Sets 6 2 2 6 2 3 Moving Window Concept for Data Storage The performance database uses a moving window concept for storing data It begins storing data in the aggregation set at the start date of the database usually the current date It stores the data records having the required date and timestamp within the date grouping TeamQuest Baseline fills in data for the specified number of date groupings When it has filled in the last date grouping it removes the data in the first date grouping When you specify the retention period for an aggregation set you are actually specifying the size of the moving window for the aggregation set In the example in Figure 6 1 TeamQuest Baseline records data at 30 minute intervals for date grouping 1 Each day it fills in another date grouping When it has filled in date grouping 61 TeamQuest Baseline removes all day 1 data when it has filled in date grouping 62 it removes all data for day 2 and so on Aggregation Set Consolidation Note The data aggregation capability is not available with TeamQuest Online The TeamQuest Baseline capability to support multiple aggregation sets in a database provides a way for consolidating data stored at a fine granularity for example a day date consolidation type at 30 minute intervals to a coarser gra
335. l Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time Logical channel write hit transfer time Logical channel write hit service time Logical channel write hit hardware service time Logical channel write hit percent of all requests IO Activity Logical Channel LCO LC1 Read Miss Logical channel read miss requests per second Logical channel read miss K words per request Logical channel read miss transfer time Logical channel read miss service time Logical channel read miss hardware service time Logical channel read miss percent of all requests IO Activity Logical Channel LCO LC1 Write Miss Logical channel write miss requests per second Logical channel write miss K words per request Logical channel write miss transfer time Logical channel write miss service time Logical channel write miss hardware service time Logical channel write miss percent of all requests IO Activity Logical Channel LCO LC1 Read Bypass Logical channel read bypass requests per second Logical channel read bypass K words per request Logical channel read bypass transfer time TQ 010
336. larm administration SNMP traps J 2 alarm handling description 2 5 7 1 introduction 3 3 user defined 7 31 alarm messages description 7 30 file of 7 27 format 7 30 viewing 7 27 alarms considerations for using TeamQuest Baseline with SPO 7 32 handling 7 28 maintaining 7 1 severity indicator 7 1 Alarms and Thresholds Menu 7 5 TQ 01011 18 alias defining 8 11 deleting 8 13 qualifying parameters 8 12 applying collection period changes 11 9 retention specification changes 11 15 B backup of performance database 10 8 C capabilities of TeamQuest Baseline 1 5 checking thresholds 7 21 Client Table Open Distributed Transaction Processing A 64 CMS 1100 configuration for SNMP traps J 2 Collection Period Menu of tqzip 11 4 collection periods administering 11 4 description 11 1 Communications Application Program Interface overview I 1 condensed output of tqrpt D 4 connections named pipe 4 6 socket 4 5 SSL 4 6 contents of SNMP traps J 4 conventions report definition file C 2 converting expanded aggregation sets 6 26 creating aggregation set 6 10 6 16 derived statistic 8 8 derived statistic threshold definitions 7 14 Index 1 Index threshold definitions 7 6 workload definition 5 13 workload set definition 5 10 D data aggregation performing 6 27 data collection control 9 1 description 1 6 1 8 of IOTRACE data 9 25 of MAPPER log data 9 29 of MIPS metering data 9 9 of Open D
337. lection To terminate the tqsam run from a demand run that has full console mode capabilities or greater enter CONS TQSAM END To terminate the tqsam run from the system console enter TQSAM END TQ 01011 18 9 4 3 9 4 4 Customizing Data Collection Probes Abnormal Termination of Data Collection If tqgsam terminates in error due to a maximum time contingency it restarts automatically If tqsam terminates and does not restart try to resolve the problem as soon as possible and restart tqsam Data is collected only when tqsam is running Workload Characteristics for the System Activity Monitor Probe Note Workload evaluation is performed each time the data is retrieved for reporting by TeamQuest Performance Software Administration This provides for dynamic workload classification as the workload definition can be changed using TeamQuest Performance Software Administration and reflected in the tqsam reports This allows you to dynamically fine tune a workload definition that might also be used in the tqsys probe The tqsam probe has a unique set of characteristics that are used for its workload definitions You use these characteristics when you create the workload definitions using TeamQuest Performance Software Administration see 5 3 The following characteristics define the tqsam workloads Name Characters Type Length Operations Run_ID A Z 0 9 string 6 Program A Z 0 9 string 12 Transaction A Z 0 9 strin
338. les by using the aggregation set granularity For example if granularity is 60 seconds and you specify a delay of 1 to 60 seconds the tqalm program will delay one sample That means that tqalm will not use the most recent data point but the point before the most recent one to do derived statistic threshold checking m is used with the r option for real time mode execution of tqalm The size parameter is used to allocate memory space before going to real time mode in tqalm The size parameter is specified in thousands of words and the default is 65 that is 65 000 words r is used to execute tqalm in real time mode A real time priority of 35 is used to run tqalm If you have problems running in real time mode you can use the m option to adjust the size of memory space allocated by tqalm before it goes to real time mode TQ 01011 18 7 29 Customizing Alarms and Thresholds 7 18 7 30 Terminating the tqalm Run To terminate the tqalm run from a demand run that has full console mode capabilities or greater enter CONS TQALM END To terminate the tqalm run from the system console enter TQALM END Alarm Messages The tqalm program generates alarm messages for each alarm that occurs on your system The alarm messages can be sent to the system console and are written to the TeamQuest Baseline alarm history table by tqalm The alarm messages displayed within a TeamQuest View session and those viewed within tqzip see 7 16
339. lied all active workload sets for the source data type hvslog are used specifies the TPM audit trail number By default this value is 17 corresponding to SYS ACI 17 and the SYS F017L1 audit trail file names used in file selection Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent line of the processor call 9 7 2 Terminating tqrtpm Data Collection To terminate the tqrtpm run from a demand run that has full console mode capabilities or greater enter the following CONS TQRTPM END To terminate the tqrtpm run from the system console enter the following TQRTPM END TQ 01011 18 Customizing Data Collection Probes 9 7 3 9 20 tqtpm Processor Call Statement The tqtpm runstream contains the tqtpm processor call statement which controls the TPM log data collection The tqtpm processor call statement specifies data collection parameters including input parameter table database name optional aggregation set name and audit trail number optional Normally you use the default values for these parameters as provided in the tqtpm runstream To modify any of these defaults you need to update the tqtpm runstream in SYS LIB RUN The tqtpm processor call has the following format tqtpm u z i parameter table L d database name a aggsetl n audit
340. ll statement tqmpr extracts statistical values on MAPPER resource utilization plus date and time information from the log file determines which workload used the resource and stores this information in the performance database according to the workloads defined for the system and the MAPPER environments Introduction 1 6 1 2 e tqsys The tqsys probe processes data from the system log files The tqsys probe automatically processes the system log files from the previous day or the system log files specified on its processor call statement The tqsys probe extracts statistical values on system resource utilization plus date and time information from the log file determines the workload that used the resource and stores this information in the performance database according to the workloads defined for the system e tqodtp The tqodtp probe collects information in an Open Distributed Transaction Processing ODTP environment The tqodtp probe retrieves information from an ODTP environment for configured clients servers and system service programs TeamQuest Baseline provides two user probe interface programs called tqusrprb and tqtblprb that allow you to develop your own site specific probes These programs coupled with user supplied information insert data into the performance database Any number of user developed probes can exist on your system The TeamQuest Probes software is also available for use with TeamQuest Baseline and
341. log primary Active 1 B system 3 C venus D saturn mars syslog secondary Active 1 E system 1 saturn F syslog primary sysjob 001 G y H Run_ID RUNNE Run_ID MSCP Run_ID CMR I Run_ID PCCTI Run_ID DDP Run_ID PXFR syslog primary cmsmcb 002 J Run_ID CMS Run_ID MCB syslog primary accounting 003 Run_ID ACC Account 1010 syslog primary venus 004 Run_ID VEN Account 2190 Account 1103 syslog secondary sysjob 001 Run_ID RUNNE Run_ID MSCP Run_ID CMR Run_ID PCCTI Run_ID DDP Run_ID PXFR syslog secondary accounting 002 Run_ID ACC Account 3333 5 20 TQ 01011 18 Customizing Workloads The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The first line of the workload definition file must be the header line BEGIN WORKLOADS A blank line must follow the header line B The first section following the header line contains a list of the workload set definitions Each workload set definition consists of the source data type the workload set name the Active Inactive flag and the number of workload set qualifiers for the workload set In this example syslog is the source data type primary is the workload set name the active flag is
342. lows any graph lines or other text that may be behind the legend to show through Border Indicates the color of the border line that appears for the background area of the legend If the Border field is not specified TeamQuest View does not display a border line around the background area of the legend The default values for Legend Style are defined in the Report Styles settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Threshold Color Normal Exception TQ 01011 18 The colors that are used to display thresholds on the report The colors indicated by these specifications appear in a small box next to the parameter names in the legend of the report Normal Is the color that appears on the report when the data value is above the lower threshold limit and below the upper threshold limit Valid colors for this field are any of the colors defined in the Threshold settings of the User Preferences Exception Is the color that appears on the report when the data value is below the lower threshold limit or above the upper threshold limit Valid colors for this field are any of the colors defined in the Threshold settings of the User Preferences The default values for Threshold Color are defined in the View Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences C 21 Report Definition File Threshold Pattern Normal
343. ls and Facility Usage Summary Using statistics derived from the tqbsp and tqsys probes this report shows overall average populations throughputs and response times for MAPPER TIP demand deadline batch and batch environments respectively The report also shows total CPU utilization disk IOs per second average disk IO response time and memory utilization allbatch rpt Overall Batch Performance Using statistics derived from the tqsys probe this report shows overall average batch environment population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per batch run alldeadl rpt Overall Deadline Performance Using statistics derived from the tqsys probe this report shows overall average deadline batch environment population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per deadline batch run alldmand rpt Overall Demand Performance Using statistics derived from the tqsys probe this report shows overall average demand environment population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per demand run Predefined Report Descriptions B 20 File Name Description File Name Description File Na
344. lt value is 132 X specifies that the report output is to be in a condensed format that is suitable for import into a spreadsheet program The delimiter parameter is c for comma delimited output or t for tab delimited output r specifies the name of a file containing the report definition If you specify the r option it must be the last parameter on the processor call 0 specifies the name of the file that receives the generated report V specifies that the output values in the report should be displayed with 5 decimal places This option only effects reports generated using the x option condensed format If the v option is not used the values in the report will be displayed with 2 decimal places All parameters are optional If a report definition file is specified the a s e c f and l options override the values specified in the report definition file If no report definition file is specified the option values are used and that item is not prompted D 2 TQ 01011 18 Host Based Reporting for Aggregation Set Data The report definitions have a fixed format The information presented in them is positionally dependent and must be in the following form aggregation_set_name report titl start_time end_time time_consolidation first_date last_date date_consolidation parameter_selectionl a parameter_selectionN The date_consolidation specifies the number of dates to consolidate for each point Th
345. lue representing the average for the parameter over a specified period of time ANSI Abbreviation for American National Standards Institute C capacity management The process of managing computer system capacity for users to ensure that system processing can and always will deliver services to meet the user s productivity goals capacity planning The function and activities concerned with determining future requirements for a computer system to meet the needs of the system users client server model A type of computer system architecture in which the components acting as the client make requests for data services or resources to those components acting as the server The server performs the necessary operations to provide the requested data services or resources to the client Communications Application Program Interface formerly called COMAPI Communications Application Program Interface is a combined extended mode subsystem and background run that allows applications to access a communications network correlation analysis The process of comparing performance statistics to analyze system data TQ 01011 18 Glossary 1 Glossary D data collection That portion of TeamQuest Baseline or TeamQuest Online host processing in which data for the performance database is collected from the data sources on the host system data display That portion of TeamQuest Baseline or TeamQuest Online processing in which TeamQuest Vie
346. lues that are shown along the vertical axis and the distance between the major and minor tick marks on the axis Major Is a real number representing the distance from the intersection of the vertical axis and the horizontal axis to the first major tick mark Major tick marks appear on every integer multiple of the real number specified in this field When you specify zero in this field the value is determined automatically Minor Is a real number representing the distance from the intersection of the vertical axis and the horizontal axis to the first minor tick mark Minor tick marks appear on every integer multiple of the real number specified in this field When you specify zero in this field the value is determined automatically Major tick marks always override minor tick marks The value specified in the Major field should be an integer multiple of the value specified in the Minor field Minimum Is a real number representing the point where the vertical axis crosses the horizontal axis Values less than the specified minimum value do not appear on the report When you specify zero in this field the value is determined automatically TQ 01011 18 C 13 Report Definition File Maximum Is a real number representing the point where the vertical axis ends on the report display Values greater than the specified maximum value do not appear on the report When you specify zero in this field the value is determined automatically
347. ly Parameter Hierarchy Category Group Category Statistic Name Locked Pages Single Page Loads Sec Queued Page Loads Sec Queued Pages Sec ACW Group Tracks Sec ACW Groups Sec Category Group Category Subcategory Statistic Name Nailed Pages Active Pages Free Pages Free Pages MIN Free Pages MAX Reserved Pages Down Pages Verify Recovering Pages Exec Pages User Pages File Pages TIP Shared Memory Paging Activity Average number of locked pages Single page loads caused by memory tight conditions pages per second Number of page load requests queued because real memory is unavailable requests per second Number of pages queued because real memory unavailable pages per second Total number of tracks written for working set output tracks per second Count of access control word ACW groups ACWs per second Memory Paging Activity Real Space Usage Pages of real memory that are nailed in pages Pages of real memory that are active in pages Pages of real memory that are free in pages Minimum pages of real memory that are free in pages Maximum pages of real memory that are free in pages Pages of real memory that are reserved in pages Pages of real memory that are down in pages Pages of real memory that are verifying or recovering in pages Pages of real memory used by Exec in pages Pages of real memory used by users in pages Pages of real memory used by files in pages
348. mQuest Performance Management Architecture host system The system where the TeamQuest Baseline or TeamQuest Online software resides IOTRACE An optional feature of SIP that can run independently or with SIP data collection It gathers detailed information about every I O operation to tape or mass storage monitor reports One of the TeamQuest View report types Monitor reports show real time performance data as the activity occurs on the system P parameter A unique performance data item in the performance database parameter group A section within a report definition file that identifies what parameter or parameters are retrieved from the host database A report definition can contain multiple parameter group sections parameter value A specific instance of data for a parameter in the performance database A parameter value is identified by the parameter name aggregation set and time and date of the point performance evaluation The analysis of system statistics to provide information on system activities and to identify any required corrective actions predefined reports The set of reports that are part of the TeamQuest View software TQ 01011 18 Glossary 3 Glossary privilege The mechanism of the OS 2200 security system that allows specified users to override mandatory and discretionary access controls probe A component of the TeamQuest Performance Management Software that collects performance data from t
349. made up of simple or compound regular expressions For example Txn_Code CSF matches on the string CSF and only CSF Txn_Code CSF 0 2 matches on the string CSFO CSF1 and CSF2 Txn_Code CSF matches on any string beginning with CSF and followed by anything Regular expressions are defined recursively with the following rules e A single character is a regular expression that matches the character itself The following characters have special significance and do not match the character itself a period a backslash brackets a hyphen a caret a plus sign parentheses or an asterisk e A period is a regular expression that matches any single character For example the period matches an A or B or any other character Run_ID SLC matches on SLC and any single character following e A backslash followed by a single character is a regular expression that matches a single character For example User_ID SLC matches on the string SLC The backslash indicates that the period is treated as a character and is included in the match e A set of characters enclosed in brackets is a regular expression that matches any one of the characters enclosed within the brackets The regular expression AB matches A or B For example Txn_Code LCSFLABC matches on CSFA or CSFB or CSFC Txn_Code CSFLXYJ0 matches on CSFX0 or CSFYO TQ 01011 18 5 5 Customizing Workloads e
350. mance Monitor TPM logging mechanism for reporting Transaction Processing TIP and High Volume Transaction Processing HVTIP statistics Single Point Autoaction Message System Administration Guide 7833 4760 This guide provides information on the Single Point Autoaction Message System SP AMS The alarm capability of TeamQuest Baseline can be integrated with the message automation and alarm handling functionality of SP AMS SMART for ClearPath OS 2200 Administration and Operations Guide 7830 7337 This guide provides information on SMART The alarm capability of TeamQuest Baseline can be integrated with the message automation and alarm handling functionality of SMART How to Access Portable Document Format PDF Files All TeamQuest product documentation is available as PDF files PDF files can either be downloaded from the TeamQuest Web site or they can be viewed from your product CD or DVD You must have the Adobe Reader software from www adobe com installed on your workstation to view the PDF files To download product documentation from the Web site Access the TeamQuest Web site at www teamquest com Click the Support tab then Customer Area Log in to the Customer Area Click the Downloads link Click the Product Documentation link for the desired release level and product platform OT AAS I 6 Select the document you wish to download When the download is complete the PDF file will open 7 Select the Save A
351. mand line 80 characters is assumed LeftHeader text A left justified heading for the report The text must be enclosed in double quotation marks Any of the special identifiers can be used in the heading text The report heading appears at the top of the page and is repeated on every page of the report CenterHeader text A centered heading for the report The text must be enclosed in double quotation marks Any of the special identifiers can be used in the heading text The report heading appears at the top of the page and is repeated on every page of the report RightHeader text A right justified heading for the report The text must be enclosed in double quotation marks Any of the special identifiers can be used in the heading text The report heading appears at the top of the page and is repeated on every page of the report LeftFooter text A left justified footer for the report The text must be enclosed in double quotation marks A footer appears at the bottom of the page and is repeated on every page of the report CenterFooter text A centered footer for the report The text string must be enclosed in double quotation marks A footer appears at the bottom of the page and is repeated on every page of the report TQ 01011 18 Host Based Reporting for Table Data RightFooter text A right justified footer for the report The text must be enclosed in double quotati
352. match The Value Types can be one of the following Average Specifies that the average value for the data sample be reported StdDev Specifies that the standard deviation value for the data sample be reported Threshold EnableLow Lower EnableUp Upper The threshold limits for the parameter value of this parameter When parameter values exceed the specified limits the color or pattern displayed in the threshold box in the report legend changes according to the specified values in the Threshold Color and Threshold Pattern statements of the general section EnableLow Indicates whether the lower threshold limit is activated for this parameter If EnableLow is set to On the threshold box appears on the report legend If EnableLow is set to Off the lower threshold is not active for this data series If both EnableLow and EnableUp are set to Off the threshold box does not appear on the report legend The default for this field is Off Lower Specifies the lower threshold limit When parameter values are less than or equal to this value the threshold box for the data series appears in the color or pattern specified in the Normal field of Threshold Color or Threshold Pattern as defined in the general section EnableUp Indicates whether the upper threshold limit is activated for this parameter If EnableUp is set to On the upper threshold appears on the report legend If EnableUp is set to Off the upper threshold is not active for this d
353. me Description File Name Description File Name Description File Name Description alltip rpt Overall TIP Performance Using statistics derived from the tqsys probe this report shows overall average TIP environment population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per TIP transaction allhvtip rpt Overall HVTIP Performance Using statistics derived from the tqtpm and tqrtpm probes this report shows overall average HVTIP environment population throughput response time CMS queue time processing time lock wait time termination time IP SUPs per second ER CC SUPs per second IO SUPs per second FCSS reads per second and FCSS writes per second allmappr rpt Overall MAPPER Performance Using statistics derived from the tqmpr probe this report shows overall average MAPPER environment population throughput response time lines processed per second and IOs processed per second batch rpt Batch Workload Performance Using statistics derived from the tqsys probe this report shows by workload batch population throughput response time IP SUPs per second IO SUPs per second ER CC SUPs per second disk IOs per second tape IOs per second and memory usage per batch run deadline rpt Deadline Workload Performance Using statistics derived from the tqsys probe this report shows
354. me Description File Name Description File Name Description TQ 01011 18 These reports are not available for TeamQuest Online WORKLOAD RPT Information by workload Gives statistics for each workload on total SUPs and transaction count OUTPUT RPT Output Messages Gives statistics for each workload on output message number and size TXNRSP RPT Transaction response time Gives statistics for each workload on total transaction response time TXNTIME RPT Transaction time Gives statistics for each workload on the components of transaction time Predefined Report Descriptions B 18 File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description FCSS RPT FCSS Function Requests Gives statistics for each workload on the number of FCSS read write lock and other requests FCSSWORD RPT FCSS Words Gives statistics for each workload on the total number of FCSS words read and written LINC RPT LINC Read Writes Gives statistics for each workload on the total number of Exec and TIP read and write requests LOCKS RPT Locks Gives statistics for each workload on the number of locks time waiting on locks and the time locks were held SUPS RPT SUPs by Workload Give statistics for e
355. me 5 minute Granularity 00 05 00 Expanded Agg Set Source Agg Set 1 minute OS ee aes 1 Week Data Collection Times Range 07 00 00 14 59 59 Type Week v al B First Day Interval 8 Hr Exclude al E Date Consolidation Aggregation Set Attributes Dialog Box The Aggregation Set Attributes dialog box contains the following fields Agg Set Name specifies the name of the aggregation set A default name is supplied by TeamQuest Baseline You can use this name or supply your own name This field is limited to 23 characters in length Granularity contains a list of the available options for the granularity of the aggregation set Granularity is the sample interval across which each data sample is taken Select the granularity for the new aggregation set from this list box 6 10 TQ 01011 18 Customizing Aggregation Sets Source Agg Set contains a list of aggregation sets already defined in the database that can be used as the source of data for the aggregation set being defined This field is used for aggregation sets that are being fed by other aggregation sets in the database If you select None the aggregation set receives data directly from the data collection probes and is not fed by another aggregation set If you select an aggregation set from the list the new aggregation set will be fed from data of the specified aggregation set The aggregation sets that appear in the Source Agg Set list box are those s
356. me sites using the Logic and Information Network Complier LINC 1100 The LINC user area format is the same as that described in 9 7 6 TQ 01011 18 9 17 Customizing Data Collection Probes 9 7 1 9 18 tqrtpm Processor Call Statement The tqrtpm runstream contains the tqrtpm processor call statement which is used to start the tqrtpm probe data collection The tqrtpm processor call statement specifies data collection parameters including input parameter table and optionally the database name aggregation set name workload set name and audit trail number To modify any of these defaults you need to update the tqrtpm runstream in SYS LIB RUN It is recommended that the tqrtpm probe be run at real time by specifying the r option otherwise the probe may fall behind while reporting data The tqrtpm probe can interpret the user area of the TPM termination records in the TPM audit trail See 9 7 6 for the LINC 1100 user formatted area that the probe can process when the u option has been specified The default is to not process the user area The tqrtpm processor call has the following format tqrtpm r u z i parameter table L d database name L a aggsetl L w workload set name L w workload set name L n audit trail number END where r executes tqrtpm in real time mode The account for the tqrtpm run must have a maximum real time level greater than zero The r option is recommended so the tqrtpm p
357. meter groups Keyname keyvalue TQ 01011 18 The definitions for the host defined keywords which identify the parameter to which the settings are applied All Keynames must be specified to match a related parameter Keyname For the Team Quest Performance Software the Keyname is one of the following keywords defined by the host system Statistic related key names Category Group Category Subcategory Statistic Note On some platforms Category and Subcategory are combined into the form Category Subcategory and called Class and Subclass in the TeamQuest View interface Qualifier related key names System Workload Set Workload Resource Application Group Unisys OS 2200 only keyvalue Can be a string of up to 52 characters which defines the Keyname The keyvalue specifications must be enclosed in quotation marks The keyvalue of signifies that this Keyname is not considered when matching the Keyname values with the actual parameter values to set the threshold and parameter settings For example if the appears as the keyvalue for the Resource then the threshold and parameter attributes will be applied to all parameters that match all the other Keyname values ignoring the Resource Keyname C 33 Report Definition File Value Types Average StdDev The specification for the type of data that is included in the report This must match the Value Types setting of the parameters you want to
358. meters IOTRACE Parameters MAPPER Log Parameters System Log Parameters Workload Performance Derived Statistics System Activity Monitor Statistics Open Distributed Transaction Processing Statistics OS 2200 Parameters A 2 SIP Parameters This subsection contains a list of the parameters collected by the SIP probe These parameters are collected by TeamQuest Baseline and TeamQuest Online e System Statistics e Processor Statistics e Dedicated Java Processor Statistics e Memory Statistics e VO Activity Statistics e Queue Statistics The SIP probe monitors the data in the category groups included in the following overview of the Category Group Category and Subcategory definitions Category Group System Category Group Processor Category Population Response Time Throughput Exec Requests Exec CALLs Category Total Processor Subcategory System Total Total System Total Total Subcategory Utilization Exec 3 Utilization Interrupts Exec Requests Switching Hardware Summary Collection Base Util by Switching Level Hardware Interrupts Collection Base Hardware Interrupts Hardware Monitor Interrupts Util by Switching Level Utilization TQ 01011 18 OS 2200 Parameters Category Group Category Processor Java Total Processor Category Group Category Memory Size EXPOOL Program Load Paging Activity Category Group Category IO Activity Physical Channel Logical Channel
359. mmunications Management System CMS 1100 Configuration Reference Manual 7830 9853 Unisys Corporation OS 2200 Communications Management System CMS 1100 Installation and Configuration Guide 7830 9846 Unisys Corporation OS 2200 Exec System Software Operations Reference Manual 7831 0281 Unisys Corporation OS 2200 Linking System Subsystems Programming Guide 7830 7451 Unisys Corporation TQ 01011 18 Bibliography 1 Bibliography OS 2200 System Log Operations and Support Reference Manual 7831 0315 Unisys Corporation Single Point Autoaction Message System Administration Guide 7833 4760 Unisys Corporation SMART for ClearPath OS 2200 Administration and Operations Guide 7830 7337 Unisys Corporation TeamQuest Model User Guide for Unisys Systems TQ 01424 TeamQuest Corporation TeamQuest View User Guide TQ 01401 TeamQuest Corporation Bibliography 2 TQ 01011 18 Index A abnormal termination of tqbsp 9 3 oftqsam 9 7 adding collection period for tables 11 7 retention specification for tables 11 13 Admin utility See TeamQuest Performance Software Administration administration sessions terminating 4 8 ageregation set compact 6 7 consolidation 6 4 converting 6 26 creating 6 10 6 16 deleting 6 15 description 1 10 2 1 displaying 6 14 modifying 6 14 multiple sets in a database 2 1 structure 6 2 tqasa administration 6 15 aggregation sets description 3 3 a
360. mp History Full Delayed Hardware Check Dayclock Performance Monitoring Initial Program Load UPI Initial UPI Normal Category Group Category Resource Subcategory Statistic Name Available Time Available Category Group Category Resource Subcategory Statistic Name Idle Total Exec Total User Misc NonSwitchable IO Processing Processor Control Memory Management TIP Integ Rec Aud Cntr Symbiont Misc Switchable Real Time TQ 01011 18 Interrupt 22 not used interrupts per second Interrupt 23 not used interrupts per second Software break interrupts per second Jump history full interrupts per second Delayed hardware check interrupts per second Dayclock interrupts per second Performance monitoring interrupts per second Initial program load interrupts per second UPI initial interrupts per second UPI normal interrupts per second Processor Processor IPO IP1 Collection Base Available time for this IP performance monitor for this sample Percent that this IP performance monitor was available for this sample Processor Processor IPO IP1 Utilization Percentage of time idle Percentage of time in Exec processing Percentage of time in user processing Percentage of miscellaneous nonswitchable Percentage of Exec I O processing Percentage of Exec processor control processing Percentage of Exec memory management processing Percentage of Exec TIP integrated reco
361. mple 1 The following example shows the most common form of an SNMP trap alarm It sends a trap containing all fields to port 162 of the host and community name configured in CMS 1100 SNMPTrap Example 2 This command limits the fields sent in the trap Only System Identifier Value and ThreshZones are included in the trap SNMPTrap Fields System Identifier Value ThreshZones Note Ifyou can format the event message derived from the trap at the management console you should probably send all trap fields The capability to send selected trap fields is provided for those situations where little or no control is available over the event message displayed on the console In those cases you may want to send only selected fields to limit message content or size TQ 01011 18 J 3 2200 SNMP Trap Alarms J 4 J 4 1 SNMP Trap Contents The contents of the SNMP trap that is sent to the management console is an SNMPv1 trap protocol data unit PDU It contains a set of fixed fields and an optional set of variable fields called variable bindings or varbinds The mapping of the trap fields is described in J 4 1 and J 4 2 SNMP Trap Mappings The following fields are present in every SNMPv1 trap PDU and are mapped as follows Enterprise This field is set to the enterprise identifier for Unisys 0S2200 TeamQuest which is 1 3 6 1 4 1 223 7 6 Agent Address This field is set to the IP address of the host on which the
362. mpts for you to select the parameter or derived statistic name for which you want to define thresholds B If more than one system exists in the database the tqzip program displays a list of the systems in the database The name DERIVEDS is selected to define thresholds for any of the derived statistics in your database See 7 8 for information on creating derived statistic thresholds At the prompt enter the system for which the threshold definition will be created You can enter either the number or the system name c The tqzip program displays a list of the statistics that are associated with the system name you chose in B At the prompt select the statistic name for which you want to create a threshold definition You can enter either the number of the statistic or the statistic name D If the statistic name you chose in C is used in multiple database parameters the tqzip program displays a list of full statistic names At the prompt select the statistic name that fully identifies the statistic for which you want to define thresholds You can enter either the number of the statistic or the statistic name 7 8 TQ 01011 18 TQ 01011 18 Customizing Alarms and Thresholds If the statistic you have selected in D is associated with a resource workload or application group the tqzip program prompts you for the required values to complete the parameter name selection This example shows the prompts for a resource name selection
363. n specification is on a separate line and is made up of four values separated by a space or a comma CollectionSpec start stop interval exclusion mask The start and stop values indicate the start and stop times of data collection and are specified as hhmm The interval is specified as mmss The exclusion mask is specified as a series of seven digits corresponding to the days of the week beginning with Sunday and ending with Saturday The mask is specified using the digit 1 to exclude a day and the digit 0 to include a day For example the mask 1000001 indicates that Sunday and Saturday are excluded from data collection If exclusion mask is not specified 1000001 is used by default A value is required for start stop and interval in each CollectionSpec The set of CollectionSpec lines in the load file replaces any CollectionSpec lines previously defined for the table If no CollectionSpec lines are specified in the load file all existing CollectionSpec lines are deleted for the table and no new ones are added The interval specification on a CollectionSpec is ignored if the table does not contain intervalized data TQ 01011 18 Customizing Tables FieldSpec Indicates whether or not a field in the table is to be collected by the TeamQuest Baseline probes Zero or more lines of FieldSpec are allowed Each FieldSpec is on a separate line and is made up of two values separated by a space or a comma FieldSpec field val
364. n the Web obtains performance data from TeamQuest Manager and displays it using a set of Java applets on a browser It is very useful for monitoring systems from a remote location and for the automated publication of management performance reports over the Web e TeamQuest View This product provides a graphical user interface for detailed data reporting and analysis TeamQuest View obtains performance data from TeamQuest Baseline TeamQuest SMFII or TeamQuest Manager for display and analysis on your workstation TeamQuest View allows you to monitor the performance of one or more systems and to quickly drill down to the exact cause of a problem The TeamQuest Performance Software family of products offers something for anyone who requires access to performance management data Figure 1 1 shows the relationship between the products within the TeamQuest Performance Software family Monitored System TeamQuest Manager TeamQuest TeamQuest TeamQuest TeamQuest View Alert On the Web Model Figure 1 1 TeamQuest Performance Software Family 1 2 TQ 01011 18 Introduction 1 2 What Can You Do with TeamQuest Performance Software The TeamQuest Performance Software family allows you to address a wide range of performance problems on your network of computers When the TeamQuest Performance Software is installed on every monitored system in your network you can deploy perf
365. n to develop the derived statistic 4 Click the Save As button to save the derived statistic in the statistic hierarchy See 8 3 4 for information on saving a derived statistic When you click New the Derived Statistic Definition dialog box is displayed 4 Derived Statistic Definition Session cll Database PRODUCTION Statistic Info PEE Save As pen Alias List a Ean E Function Formula 8 8 TQ 01011 18 Customizing Derived Statistics Derived Statistic Definition Dialog Box The Derived Statistic Definition dialog box contains the following information about the session and database for which you are creating a derived statistic definition Statistic Info supplies a description of the derived statistic This information is saved in the database and is available for viewing when subsequently opening a derived statistic definition This field is limited to 256 characters in length Alias List provides a list of reference names to other parameters in your database These reference names can then be used in the formula of a derived statistic The alias list is maintained by using the Define Qualify View and Delete buttons to the left of the edit box A separate independent alias exists for each derived statistic definition Define button allows you to add a statistic reference to the alias list see 8 3 3 1 Qualify button allows you to qualify the statistic referenced by the alias name by specifyin
366. nce statements of the general section TQ 01011 18 C 35 Appendix D Host Based Reporting for Aggregation Set Data TeamQuest Baseline provides the capability to generate reports from the host system without using the TeamQuest View interface The tqrpt program can generate a report based on a report definition stored in a file or can interactively solicit the information required to generate a report The printed reports are normally 132 characters wide with a separate column for each parameter reported Note The tqrpt program is not available for TeamQuest Online The processor call has the following format tqrpt d databasename a aggset where d a C TQ 01011 18 s starttime e endtime c time consolidation f firstdate 1 lastdate p w pagewidth x delimiter r report 0 filename v specifies the database from which you want to generate a report The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 specifies the aggregation set to use for retrieving data The aggset parameter is the name of an existing aggregation set If no aggregation set is specified tqrpt uses the aggregation set with the finest granularity specifies the start time for the data The starttime parameter is in the form of hhmmss The default is 000000
367. nd recovering the system This manual also contains information on the Software Instrumentation Package SIP TeamQuest Baseline and TeamQuest Online use data collected by SIP for data reduction and report generation ClearPath OS 2200 Software Products Installation Guide 7831 0612 This guide explains how to use the Software Library Administrator SOLAR to install OS 2200 software products from a software package tape OS 2200 Exec System Software Operations Reference Manual 7831 0281 This manual explains the operator interface with the system console and peripherals It also discusses operator responsibilities describes procedures and lists keyins OS 2200 Linking System Subsystems Programming Guide 7830 7451 This guide describes software subsystems including how to create install and use them It includes a discussion of security as it relates to the installation and usage of subsystems OS 2200 System Log Operations and Support Reference Manual 7831 0315 This manual explains system log processing and the resulting log entries of the OS 2200 system The TeamQuest Baseline system log probe collects system log data for its reporting viii TQ 01011 18 About This Guide OS 2200 Transaction Processing Administration and Operations Reference Manual 7830 7881 This manual tells you how to turn on the Transaction Processing TIP log accounting feature if you have TIP running on your system It also describes the TIP Perfor
368. ndicates whether the data is delimited by commas The acceptable values are ON or OFF ON indicates that the data is delimited by commas OFF indicates that the data is delimited by tabs Auto Multiplier ON OFF Determines if the Auto Multiplier feature is turned on or off ON indicates the Auto Multiplier feature is turned on OFF indicates the Auto Multiplier feature is off The default value for Auto Multiplier is defined in the View Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Display Markers ON OFF Determines if display markers appear on the report ON indicates that the display markers are turned on OFF indicates that the display markers are turned off The default value for Display Markers is defined in the View Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Sequence Color Hatch Marker Line TQ 01011 18 The attributes of the graphic representations are inherited from the multiple sequence statements that appear in the report definition The order the attributes uses is determined by the order of the sequence statement in the report definition Color Specifies the color that is used for a particular graphic object All valid colors are defined in Sequences settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Hatch Specif
369. ne or two aggregation sets directly populated by probes Other aggregation sets must be fed from those aggregation sets and filled by tqagg Aggsetl 1 minute Aggset2 10 minute 3 2 TQ 01011 18 3 2 3 3 TeamQuest Baseline Administration Aggregation Sets The TeamQuest Baseline capability to support multiple aggregation sets in a single database provides a way for consolidating data stored at a fine sampling rate By default the performance database released with TeamQuest Baseline has four aggregation sets The first aggregation set stores data sampled at 1 minute intervals for a period of 8 hours The second aggregation set holds data sampled at 10 minute intervals for a period of 8 days The third aggregation set holds data sampled at 1 hour intervals for a period of 5 weeks The fourth aggregation set holds data sampled at 8 hour intervals for a period of 13 months By default the TeamQuest Baseline probes store data into the 1 minute and the 10 minute aggregation sets The aggregation process carried out by the tqagg program in turn consolidates the 10 minute data into the 1 hour aggregation set and the consolidated 1 hour data into the 8 hour aggregation set See Section 6 for information on customizing the aggregation sets for your data collection Notes e All of the default aggregation sets are in compact format If you prefer expanded aggregation sets you must use either TeamQuest Performance Software Administration
370. ne and TeamQuest Online User Guide contains information specific to the host component of the software The TeamQuest View User Guide describes the workstation software and explains how to use the TeamQuest View user interface to view reports from data collected by the TeamQuest Performance Software applications This guide is divided into two parts Part 1 contains information that will familiarize you with the TeamQuest Baseline product and the basic concepts needed to use and administer TeamQuest Baseline as installed It is intended for first time users and those who have no need to change the default settings Part 2 contains advanced concepts and information for customizing TeamQuest Baseline to meet the specific needs for your site Organization This guide contains the following Part 1 Getting Up and Running Section 1 Introduction This section provides an introduction to TeamQuest Performance Software family It describes the TeamQuest Baseline and TeamQuest Online products and tells how the products work Section 2 TeamQuest Baseline and TeamQuest Online Concepts This section explains the architecture of TeamQuest Baseline and TeamQuest Online and describes the performance database structure Section 3 TeamQuest Baseline Administration This section provides basic information on administering TeamQuest Baseline or TeamQuest Online with the default settings as installed It contains information on default aggregation sets
371. ne tqusrprb p m L n cycle x seconds L r k keyin name L d databasename L s system g catgroup c category b subcat L a aggsetl h END where p specifies that the timestamps supplied in the PRBDATA file represent point in time timestamps and data is stored within the corresponding aggset interval Without the p option data is stored one interval earlier to match data collected by other probes such as tqbsp m specifies that tqusrprb is to run continuously n cycle specifies the cycle of PRBDATA to start processing If not supplied the current cycle of PRBDATA is used x seconds specifies the time in seconds to wait for the next cycle of PRBDATA ifit is not yet availalbe The default value is 5 seconds r specifies to run tqusrprb in real time TQ 01011 18 k keyin name d databasename 8 system g catgroup c category b subcat a aggsetl END Customizing Data Collection Probes specifies the keyin name that is registered and used for terminating the tqusrprb run The default keyin name is TQUSR specifies the database into which data is stored The databasename parameter is the name of an existing database on your system If the d option is not specified the Database setting in the tq ini element is used see 3 1 specifies the system key name in the parameter hierarchy of the statistic being stored If the system is not specified the name of th
372. ng derived statistic threshold definition Select View Thresholds on the Alarms and Thresholds Menu see 7 3 by entering 2 at the prompt The tqzip program then prompts you with a list of parameter and derived statistic names that currently have threshold definitions in the database as shown in the following example gt Enter your selection gt 2 gt gt Current Threshold Definitions A gt 1 DERIVED Derived Derived Category Derived Subcategory UserCpuUtil gt 2 DERIVED Derived Derived Category Derived Subcategory RespTimeGoal gt 3 CLO1 10 Activity Disk I0s Sec gt 4 CLO1 10 Activity Disk Summary Percent Busy DSKF1 gt gt Enter number name lt blank gt for more or E to exit gt gt gt gt DERIVED Derived Derived Category Derived Subcategory UserCpuUtil B gt 1 Threshold checking ON gt 2 Threshold 1 00075 00 Threshold 2 00000 00 gt 3 Number of zones 22 Normal zone tod gt 4 Secs before alarm 0000 Secs delay 0000 gt 5 Alarm identifier UserCpuUtil gt 6 User text Call John at ext 2332 gt 7 User text gt 8 User text gt gt List System Names lt y gt n C gt y gt gt venus gt gt List Resource Names lt y gt n gt y gt gt IP000 gt IP001 gt gt Do you want to view another threshold y lt n gt D gt N TQ 01011 18 Customizing Alarms and Thresholds The following lettered descriptions
373. ng on those parameters that have threshold definitions The probe first determines the zone in which the current value of the parameter is found The zone is determined as follows threshold value 1 is t1 threshold value 2 is t2 2 Zones Defined Zone 1 Zone 2 t1 Values lt t1 are in Zone 1 Values gt or t1 are in Zone 2 3 Zones Defined Zone 1 Zone 2 Zone 3 t1 t2 If t1 lt t2 Values gt or t2 are in Zone 3 Values gt or t1 and lt t2 are in Zone 2 Values lt t1 are in Zone 1 If t1 gt t2 Values gt or t1 are in Zone 1 Values lt t1 and gt or t2 are in Zone 2 Values lt t2 are in Zone 3 Figure 7 1 Threshold Zones TQ 01011 18 7 3 Customizing Alarms and Thresholds The following example illustrates the usage of alarms and thresholds on a system You might want a message displayed on the console if demand response time exceeds 5 seconds for a duration of at least 60 seconds You would define two zones where Zone 1 is the normal zone and threshold value 1 t1 has a value of 5 seconds The following situations could occur If demand response time is 3 seconds the parameter is in Zone 1 and no alarm is issued If demand response time increases to 10 seconds the parameter moves to Zone 2 then remains at this level for 40 seconds and decreases to 3 seconds the parameter moves to Zone 1 no alarm is issued If demand response time increases to
374. ng tqzip you can generate a file containing a parameter list using the o option You can edit this file to contain only the parameters that you want to delete from your database Then use the p option on the processor call to delete the parameters specified in the parameter file When you delete a parameter from the database all of the data values for that parameter are cleared in each of the aggregation sets in the database All space previously used by deleted parameters is automatically reused when new parameters are created Also any threshold definitions are removed for the parameters being deleted Disk space is not released when a parameter is deleted Note Parameter deletion should be a controlled operation and is best performed after you have terminated all TeamQuest Baseline probes executing on your system Be careful that you do not delete any parameters that are being used by a probe that is currently executing on your system Command Line Format To remove parameters from your database using tqzip use the following processor call format tqzip d databasename p parm file where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 p deletes parameters from the database using the file specified by the parm file parameter The parm file
375. nique user ids for each user s TQTCP run and controls which users can run TQ View or TQ Administration Configuration Element You supply information in the TQTCP configuration element SYS LIB TQ BASELINE TQTCP CONFIG that is used by the master TQTCP run to start subsequent TQTCP runs associated with each new socket connection When a user connects using a socket through TQ View or TQ Administration a new TQTCP run is initiated for each session Each statement in the TQTCP configuration element contains a label and a value separated by a space The can be used to indicate a comment line There are two different formats of information that can be entered in the TQTCP configuration element If you prefer that all TQTCP runs are started with the same user id password account and so on see I 2 3 1 If you prefer that each TQTCP run be identifiable to the user id that started it see I 2 3 2 Using this format enabled by using the u option on the TQTCP call line will restrict which user s can use TQ View or TQ Administration in accessing the TeamQuest Performance database Both TQ View and TQ Administration solicits a user id and password to be used in initiating the associated TQTCP run If the user id entered exists in the TQTCP configuration element then that user is allowed access The TQTCP run will be successfully started if the user id password and account are valid in signing on to the 2200 system When the tqinit is p
376. nment of the text within the best fit annotation box The acceptable values are Top Aligns the text at the top of the best fit annotation box Center Centers the text vertically in the best fit annotation box Bottom Aligns the text at the bottom of the best fit annotation box TQ 01011 18 C233 Report Definition File C 24 Horizontal Font Style Size Defines the justification of the text within the best fit annotation box The acceptable values are Left Justifies the text to the left in the best fit annotation box Center Centers the text horizontally in the best fit annotation box Right Justifies the text to the right in the best fit annotation box Specifies the typeface in which the best fit annotation text is printed on the report Specifies the kind of emphasis for the typeface of the best fit annotation text Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the text However the styles must be joined by a plus sign For example to specify that the best fit annotation text be printed in bold and italic typeface you enter bold italic in the Style field of the specification Specifies the height of the characters measured in points of the typeface in which the best fit annotation text is printed The value must be a positive integer Foreground Indicates the color in which the text of the best fit
377. ns that use account numbers in the range of 100030 through 100039 100120 through 100129 100150 through 100159 or 100179 Account 10003 0 9 Account 10012 0 9 Account 10015 0 9 Account 100179 User Data Probes Note The user data probes capability is available with TeamQuest Baseline only Team Quest Baseline provides the capability to create simple user probes to collect and store data into the performance database This capability is useful if there are statistics that you would like to gather but that are not collected by any of the system or application probes supplied with TeamQuest Baseline To make a simple user probe of your own write an addstream that executes the various programs that it takes to create the input needed for the TeamQuest Baseline user probe programs This may be SSG skeletons Interactive Processing Facility IPF procedures or a high level language program to collect the desired statistics You then use the TeamQuest Baseline user probe programs to read the data and store that data in the performance database The addstream must provide the data in the format expected by the TeamQuest Baseline user probe programs TQ 01011 18 Customizing Data Collection Probes The user probe programs that you can use when writing your own simple probe are called tqusrprb and tqtblprb The one you use depends on the type of data you are collecting e Use the tqusrprb program when collecting performan
378. nsidered as the ending times of the interval and tqusrprb stores the data with a timestamp that is one interval before the given DATE and TIME This is done to be consistent with other TeamQuest Baseline probes so that the user probe data correlates correctly with the data collected by the other probes For example if some data is fed as input to tqusrprb as follows 01 14 11 16 31 00 36855 11905 21270 36 areal 01 14 11 16 31 00 249991 206490 18511 92 area2 01 14 11 16 31 00 125247 19166 93561 17 area3 01 14 11 16 31 00 134839 104380 16979 86 area4 It will be stored into the 16 30 00 point reported by the 10 minute aggregation set If you prefer to store data within the same interval of the DATE and TIME you provide you may specify the p option in the tqusrprb call statement Note The tqusrprb program stores the last group of samples within any aggregation set interval being used For example if the PRBDATA file contains 30 second samples and you are storing into a 1 minute aggregation set only the last 30 second sample within the 1 minute aggregation set interval is stored In this example your program that provides the samples in the PRBDATA file must do the combining of the 30 second samples into 1 minute samples and then write the entire 1 minute sample to PRBDATA file so the tqusrprb stores the entire 1 minute data values 18 9 45 Customizing Data Collection Probes 9 11 1 2 Running tqusrprb in Continuous Mode 9 46
379. nt interrupts per second Quantum timer interrupts per second Interrupt 21 not used interrupts per second Interrupt 22 not used interrupts per second Interrupt 23 not used interrupts per second Software break interrupts per second Jump history full interrupts per second A 15 OS 2200 Parameters A 16 Delayed Hardware Check Dayclock Performance Monitoring Initial Program Load UPI Initial UPI Normal Category Group Category Resource Subcategory Statistic Name Available Time Available Category Group Category Resource Subcategory Statistic Name Idle Total Exec Total User Misc NonSwitchable IO Processing Processor Control Memory Management TIP Integ Rec Aud Cntrl Symbiont Misc Switchable Real Time ILES CMS 1100 UDS TIP Delayed hardware check interrupts per second Dayclock interrupts per second Performance monitoring interrupts per second Initial program load interrupts per second UPI initial interrupts per second UPI normal interrupts per second Processor Java Processor IPO IPI Collection Base Available time for this IP performance monitor for this sample Percent that this IP performance monitor was available for this sample Processor Java Processor IPO IP1 Utilization Percentage of time idle Percentage of time in Exec processing Percentage of time in user processing Percentage of miscellaneous nonswitchable Percentage of Ex
380. ntain the field names for the statistics to be stored The field names can appear in any order and must be separated by at least one blank space The maximum length of this line is 255 characters The following field names are allowed DATE When a field is designated DATE the value of the corresponding field in the data sample line see 9 11 1 1 is used for the date stored The date is in the format mm dd yy If none of the fields in the first line of input is designated as a DATE field the current date is used for all data samples stored When you supply a DATE field you must also supply a TIME field TIME When a field is designated TIME the value of the corresponding field in the data sample line see 9 11 1 1 is used for the time stored The time is in the format hh mm ss If none of the fields in the first line of input is designated as a TIME field the current time is used for all data samples stored statistic name The statistic name is the name of any statistic you want to collect on your system You can enter any number of statistic names Statistic names containing spaces must be enclosed in double quotation marks Some reserved words cannot be used as a statistic name because they have special meaning when specified on the first line Do not use NONE DATE TIME and RESOURCE as a statistic name TQ 01011 18 9 41 Customizing Data Collection Probes 9 42 RESOURCE When a field is designated RESOURCE the val
381. ntion specification the order of the retention specifications is assumed to be in increasing interval size from the primary retention up to the largest interval Note Tables containing nonintervalized data such as the Alarm Log table cannot have additional retention specifications because the data records cannot be aggregated across time Select Add a Retention Specification from the Retention Specification Menus see 11 6 by entering 3 at the prompt The following prompts are displayed gt Enter your selection lt 1 gt gt 3 gt gt Minimum Interval hh mm ss A gt 15 00 gt Retention period i e 3 weeks B gt 1 week The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A Enter the interval for the data records B Enter the retention period You must enter a number and a unit Valid units are minute hour day week month and year In the example a retention specification is added to the existing set of retentions The interval of the retained records is 15 minutes and the data is retained for 1 week When you add this retention to the existing set the retentions now appear as gt gt Retention Specifications 5 gt gt Number Interval Keep gt witiwoe eS eee ee Sw toe es gt Primary 9 1 day gt 1 00 01 00 1 week gt 2 00 15 00 1 week gt 3 00 30 00 1 month gt 4 01 00 00
382. nularity a week date consolidation type at 1 hour intervals The aggregation sets can reduce the amount of data that is retained in the database at the expense of losing some of the detail of the data as it was originally created When data is consolidated from one aggregation set into another aggregation set the consolidation can occur in two dimensions e Time consolidation You can consolidate your data using time consolidation For example you might have data that is collected at 15 minute intervals and wish to consolidate this data to a granularity of 60 minutes The original aggregation set has a granularity of 15 minutes You can create a second aggregation set that has a granularity of 60 minutes The data from the 15 minute aggregation set is used as the source for the 60 minute aggregation set Each of the four 15 minute points from the 15 minute aggregation set are consolidated into one point of data for the 60 minute aggregation set the four data points for 10 00 to 10 59 in the original aggregation set are consolidated into one data point for 10 00 to 10 59 in the 60 minute aggregation set The 15 minute aggregation set is the source of the data for the 60 minute aggregation set The 60 minute aggregation set is said to be fed by the 15 minute aggregation set When you define an aggregation set you can optionally specify the aggregation set that is the source of the data that is the aggregation set that feeds the newly created ag
383. o use the new aggregation sets you will need to update the runstreams or the tq ini element However sometimes you might create an aggregation set that is fed by another aggregation set This aggregation set will automatically be handled by the tqagg program without requiring any updates to the runstreams Also an aggregation set may be created only to load data from another system In that case you do not want to update the probe runstreams in SYS LIB RUN or the tq ini element TQ 01011 18 6 15 Customizing Aggregation Sets 6 4 1 6 16 Creating an Aggregation Set You can use the tqasa program to create an aggregation set for a database When you create an aggregation set a new database file is created and the tqasa program displays the name of the file Note When you create an aggregation set that contains data consolidated from another aggregation set you must run the tqagg program to consolidate this data see 6 4 7 To create an aggregation set the tqasa processor call has the following format tqasa 7 aggset d databasename g granularity s starttime e endtime t grouptype k keepcount f fed by x exclude w weekstart p where i specifies the aggregation set to create in the database The aggset parameter is the name of the new aggregation set d specifies the database for which you want to create an aggregation set The databasename parameter is the name of a database t
384. ocesses the set of IOTRACE files specified on the tqiot program call line The tqiot program should run each day that you collect IOTRACE data to ensure that your performance database contains data for each day TQ 01011 18 9 25 Customizing Data Collection Probes 9 8 2 tqiot Processor Call Statement The tqiot runstream contains the tqiot processor call statement which controls the IOTRACE data reduction The tqiot processor call statement specifies data collection parameters including input parameter table name database name aggregation set name IOTRACE file names and system name Normally you use the default values for these parameters as provided in the tqiot runstream To modify any of the defaults you need to update the tqiot runstream in SYS LIB TQ BASELINE The tqiot processor call has the following format tqiot z i parameter table L d database name L a aggsetl f IOTRACE log n L f 2OTRACE log ntl S system name END where Z is used to register the tqiot probe in the database i specifies the location of the input parameter table name to be used by the tqiot probe For the tqiot probe the parameter table is the element SYS LIB TQ BASELINE IOTINPUTTBL This is a required parameter on the tqiot call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not spec
385. ocket connections The port number defaults to the value 2700 If you need to remove any host names or IP addresses select the item you wish to remove from the list and press your keyboard Delete key TQ 01011 18 4 5 TeamQuest Performance Software Administration 4 4 2 4 4 3 4 5 4 6 Using a SSL Connection When you select the SSL option button from the New Session dialog box you need to specify a host name or an Internet Protocol IP address and a port number You can either enter a host name or IP address in the edit box or you can choose from the current list of hosts below the edit box that are already defined as part of a session You may need to specify the full domain name of the host to which you want to connect Each time you define a new session the host name or IP address is added to the list of currently defined hosts Note Before using sockets you must make sure your host software is configured to accept socket connections on that host See Appendix I for more information on socket connections The port number defaults to the value 2703 If you need to remove any host names or IP addresses select the item you wish to remove from the list and press your keyboard Delete key Using a Named Pipe Connection When you select the Named Pipe option button from the New Session dialog box you need to specify a host name or IP address You can either enter a host name or IP address in the edit box or you can c
386. odtp Client Table Information The Client table contains information specific to the ODTP environment This information is reported in a tabular column format and is displayed by TeamQuest View See A 10 3 for a list of the information stored in the Client table by tqodtp Service Table Information The Service table contains information specific to the ODTP environment This information is reported in a tabular column format and is displayed by TeamQuest View See A 10 4 for a list of the information stored in the Service table by tqodtp TQ 01011 18 Customizing Data Collection Probes 9 7 Real Time TPM and TPM Log Data Collection Note The real time TPM RTPM and TPM log probes are released with TeamQuest Baseline only These probes are not available for TeamQuest Online The tqrtpm and tqtpm log probes report the TIP HVTIP statistics from the TIP Performance Monitor TPM audit trail The tqrtpm probe runs continuously to report the data in a near real time mode The tqtpm probe reports data from the previous day while the tqrtpm probe reads the TPM audit trail as new data is written Both probes can be used on your system The tqrtpm probe reports data into one aggregation set to be viewed continuously and provides the ability to view the TIP HVTIP transactions Normally the tqrtpm probe reports 1 minute data into the 1 minute aggregation set which is not aggregated by tqagg The tqtpm probe can be configured to start after midn
387. of the fields Updating Derived Statistic Threshold Definitions The tqzip program allows you to update threshold definitions that are currently defined in the performance database To update a derived statistic threshold definition select Define Thresholds on the Alarms and Threshold Menu see 7 3 by entering 1 at the prompt The tqzip program displays the following prompts to update thresholds for a derived statistic gt Enter your selection gt 1 gt gt Create threshold Update threshold or Exit c u lt e gt A gt u gt gt Current Threshold Definitions DERIVED Derived Derived Category Derived Subcategory UserCpuUtil Vv gt 3 CLO1 10 Activity Disk I0s Sec gt 4 CLO1 10 Activity Disk Summary Percent Busy DSKF1l gt gt Enter number name lt blank gt for more or E to exit gt gt gt Select the resource s to use for this derived statistic gt TQ 01011 18 gt 2 DERIVED Derived Derived Category Derived Subcategory ResplTimeGoal Customizing Alarms and Thresholds gt Resource Name precede choice with to deselect gt 1 IPO00 gt 2 IPOO1 gt 3 TP002 gt 4 TP003 gt gt Enter number name lt blank gt for more or E to exit D gt 2 gt gt Resource Name precede choice with to deselect gt 1 IPO00 gt 2 IP001 gt 3 TP002 gt 4 TP003 gt gt Enter number name lt blank gt for more or E to exit E gt E T
388. ogram command format E 2 default reports E 12 description E 1 exporting data E 12 report definition file E 3 tqrorg utility command line format 11 28 description 11 26 reorganizing table files 11 26 tqrpt program condensed output D 4 description D 1 reporting derived statistics D 5 tqrtpm program description 1 9 processor call statement 9 18 terminating data collection 9 19 tqsam program abnormal termination 9 7 description 1 9 9 5 parameters collected by A 58 processor call statement 9 5 table data 9 8 workload characteristics 9 7 tqsys program description 1 10 9 33 parameters collected by A 46 A 54 processor call statement 9 34 system log file input 9 35 workload characteristics 9 37 Index 8 tqtblprb probe collecting and storing data 9 62 command line format 9 62 creating atable 9 62 data fields 9 49 description 1 10 9 48 example for collecting and storing data 9 58 field definition 9 49 field definition guidelines 9 49 group key fields 9 49 input format 9 48 primary key fields 9 49 removing atable 9 62 table data value formats 9 65 table definition 9 49 table specification 9 49 writing a probe for table data 9 48 TQTCP executable background run I 3 configuration element I 4 description I 2 errors I 9 processor call 4 security considerations I 10 using encryption I 10 tqtpm program description 1 9 9 17 log file input 9 21 parameters collected by A 41 processor c
389. ogram allows you to extract statistical data from the performance database in the format required by TeamQuest Model With the tqgetm program you select the database workload set system and the time and date period that you want to model tqgetm retrieves the required data from the performance database and writes the data to a file that can be used as input to TeamQuest Model See the TeamQuest Model User Guide for information on how to use the tqgetm program Note The tqgetm program requires a special access license key to operate Contact TeamQuest Corporation to receive a key TQ 01011 18 Appendix G Alarm History Table Alarms raised due to threshold violations are handled by the tqalm program The tqalm program writes each alarm to the alarm history table The tqzip program can be used to export alarm messages that were generated within a specified time interval The alarm history table is automatically created by the tqalm program The alarm history table is created in file TQDATA xxxxalarmlog The xxxx is the first four characters of the database name for which the alarm history table is being created The following are details of the alarm history table Table Field Hierarchy Class Alarm Name Alarm Log File name samfile Collection interval 00 00 00 23 59 59 Default retention 1 month Data Item Type Timestamp Integer Sequence_Number Integer System String Severity String Alarm_ID String Qualifier_1
390. olumn To indicate a relative position the number should be preceded by a plus sign The relative position is calculated from the position where the previous character was printed on the page Optionally you can indicate just a row specification by row In this case the text is centered on the page You can specify more than one TextField TQ 01011 18 Host Based Reporting for Table Data DataFields fieldname format row col text TQ 01011 18 Specifies how the data from the table is displayed in the report The fieldname must be a valid field name for the table The format specifies how the data for the specified field is displayed The format is optional and must be enclosed in double quotation marks It can be one of the following LJn The data is left justified and n characters are displayed RJn The data is right justified and n characters are displayed YY The year is displayed as a 2 digit number YYYY The year is displayed as a 4 digit number MM The month is displayed as a 2 digit number MMM The month is displayed as a 3 character string DD The day of the month is displayed as a 2 digit number hh The hour is displayed as a 2 digit number mm The minutes are displayed as a 2 digit number ss The seconds are displayed as a 2 digit number The data is displayed as a decimal number where each number sign represents one digit The data is displayed as a floating point number wher
391. om the workload definition file without first deleting the existing definitions in the database If you do not specify the e option all existing definitions are deleted from your database before the new ones are loaded TQ 01011 18 Example Customizing Workloads The following shows an example of loading workload definitions from the file wfile TQ 01011 18 gt tqzip l w wfile gt TQZIP 7R4G 011312 1104 55 Fri Jan 13 14 01 11 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt gt Using database TQDATA PRODUCTION gt Work gt Wor gt Wor gt Wor gt Wor gt Wor gt Wor gt gt Work gt K K K K K K gt End T Oad oad Oad Oad Oad Oad Oad Oad QZIP Set workloadsetl has been inserted modified Set workloadset2 has been inserted modified workloadl has been inserted modified workload2 has been inserted modified workloadl has been inserted modified workload2 has been inserted modified load complete definitions loaded 5 23 Section 6 Customizing Aggregation Sets 6 1 6 2 Introduction TeamQuest Performance Software Administration or the tqasa program can be used to administer the aggregation sets in your performance database Default aggregation sets are created during installation If you are content with the default aggregation sets no additional administration is required However if you want to customize your aggr
392. on TQ 01011 18 C 5 Report Definition File Point Size size Specifies the size of one data point in seconds Size can be any real integer If Point Size 3600 the granularity of the data in the aggregation set defines the time span of the data points graphed For example if the granularity of the data in the aggregation set is 30 seconds then 30 second data points will be retrieved A value of 3600 indicates one point per hour If the granularity of the data is 30 minutes then two data points are consolidated to create a single point Time start end The time interval for which data is requested from the host These values are expressed in a 24 hour clock notation start Is the starting time for the time interval in the format hh mm ss end Is the ending time for the time interval in the format hh mm ss The time interval cannot be greater than 24 hours If the ending time is less than or equal to the starting time TeamQuest View assumes that the time interval extends across midnight to the next day If the report is a monitor the difference between the start time and the end time determines the size of the report that is displayed Date Calendar Type Relative Date Relative Date The relative date for the default calendar CalendarType Is the Day Week Month Quarter or Year that determines the relative calendar type and the representation of RelativeDate If CalendarType is not present CalendarType is
393. on e specifies the month day and year of the ending date used in the log file selection n specifies which audit trail number to use This option should only be used to read MIPS data from a trail other than the dedicated COD audit trail To read from the system log file use n audit trail number specification where audit trail number is the system log trail number Processing from the COD audit trail is much more efficient than using the n option TQ 01011 18 9 11 Customizing Data Collection Probes 9 5 5 9 12 If a 2 digit year is specified the digits 00 through 27 are interpreted as the years 2000 through 2027 and the digits 28 through 99 are interpreted as the years 1928 through 1999 You can also set a relative time option when selecting the log entries to process The format of the relative time option is t relative hour L n audit trail number where t specifies a relative hour when selecting log entries to process The valid range is 1 to 24 For example t 8 selects and processes the last 8 hours of log data relative to when the tqtmips program was started Note A value greater than 24 selects and processes all of the previous days data n specifies which audit trail number to use This option should only be used to read MIPS data from a trail other than the dedicated COD audit trail To read from the system log file use n audit trail number specification where audit trail number is the system lo
394. on Data Exception Operation Trap Breakpoint Quantum Timer Interrupt 21 Percentage of TIP processing Minimum percentage of TIP processing Maximum percentage of TIP processing Percentage of critical deadline processing Percentage of demand processing Percentage of batch processing Software Controlled Performance level factor for the system Scaled System controlled Performance level factor for the system Contracted performance level percent for the system Processor Total Hardware Interrupts Hardware default interrupts per second Unretryable hardware check interrupts per second Diagnostic interrupts per second Interrupt 3 not used per second Interrupt 4 not used per second Interrupt 5 not used per second Interrupt 6 not used per second Interrupt 7 not used per second Reference violation interrupts per second Addressing interrupts Terminal addressing exception interrupts per second RCS generic stack underflow overflow interrupts per second Signal interrupts per second Test and set interrupts per second Invalid instruction interrupts per second Page exception interrupts per second Arithmetic exception interrupts per second Data exception interrupts per second Operation trap interrupts per second Breakpoint interrupts per second Quantum timer interrupts per second Interrupt 21 not used interrupts per second TQ 01011 18 OS 2200 Parameters Interrupt 22 Interrupt 23 Software Break Ju
395. on contained in the load file To load table specifications with tqzip use the following command line format tqzip d databasename 1 c tablespec file TQ 01011 18 11 25 Customizing Tables where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting of the tq ini element is used see 3 1 l specifies that a load operation is to be performed by tqzip The type of load operation is table collection retention as indicated by the c option C loads the data collection and retention specifications into the performance database from the file specified by the tablespec file parameter The tablespec file parameter is the name of a data and retention specification file created by the u and c options of tqzip see 11 10 Example The following shows an example of loading table specifications from the file tablefile to the performance database gt tqzip l c tablefile gt TQZIP 7R4G 011312 0704 45 Fri Jan 13 08 15 27 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt Using database TQDATA PRODUCTION gt gt Table information loaded gt gt End TQZIP 11 12 Reorganizing Table Files 11 26 Table data such as process data and alarm logging is maintained in a set of files in the performance database The tqrorg utility is pr
396. on marks A footer appears at the bottom of the page and is repeated on every page of the report ReportTitle row col text A title for the report that is printed on the first page of the report at the specified row and column The row col indicates the row and column on the page where the text string should start The parentheses must be included in the row col specification You can specify the row or column number in one of two ways To indicate the absolute row or column specify only a number for the row and column To indicate a relative position the number should be preceded by a plus sign The relative position is calculated from the position where the previous character was printed on the page Optionally you can indicate just a row specification by row In this case the text is centered on the page The text must be enclosed in double quotation marks You can specify more than one ReportTitle key ColumnHeadings YES NO Determines if column headings appear on all pages of the report If set to YES the column headings appear on all pages of the report If set to NO the column headings are printed on the first page of the report only The report heading and footer continue to be printed on each page Delimiter COMMA SPACE TAB text The delimiter to be used between columns in the report The delimiter and be one of the reserved words TAB SPACE or COMMA or it can be any text string
397. one character These keywords are used to define an exclude mask pattern It is used to exclude choices from a list based on a mask statement or an include mask pattern The patterns can be made up of any character with the following mask characters available Matches zero of more choices Matches exactly one character TQ 01011 18 Report Definition File parameter group This keyword allows you to specify the name of a previous parameter group to use the keyword values defined for that group The referenced parameter group must be a previously defined parameter group Value Types Average StdDev All The specification for the type of data that is included in the report Each data sample in the database contains the average and standard deviation values for the specified time interval You must enter one of the following choices in this field Average Specifies that the average value for the data sample be reported StdDev Specifies that the standard deviation value for the data sample be reported All Specifies that all values for the data sample be reported The default value for Value Types is Average Top Positive integer StatisticItem LastPoint The specifications for selecting the most active resources for a given statistic Positive integer Specifies the number of the most active resources that are displayed A can be specified instead of an integer value to allow you to sele
398. onfiguration element interface CPCOMM port 2700 userid tquser password 123456 runid tqr account 11000 project tqproj siteid TQ001 siteid TQ002 siteid TQ003 siteid TQ004 siteid TQ005 xqtfile tq file Note Remember that the only required statements are userid and account If the defaults for the other configuration statements are acceptable you do not need to include any of the optional configuration statements in your TQTCP configuration element In this example TQTCP would use the CPCOMM interface to listen on port 2700 for incoming connections from TeamQuest View or TeamQuest Performance Software Administration When a connection is made an RSI terminal would be initiated using one of the five specified site ids TQ001 through TQ005 The RSI run would be started using the following user id password and RUN statement tquser 123456 run tqrun 11000 tqproj The following processor call statement would be used to execute tqdata in the RSI run socketid is inserted by TQTCP tq file tqdata c socketid 6 TQ 01011 18 2200 Socket Interface 1 2 3 2 Since five site ids were specified in this example TQTCP would allow up to five connections operating concurrently and would use the specified site ids TQ001 through TQ005 when initiating RSI runs to handle each connection If no site id statements were included in the TQTCP configuration element up to 50 concurrent connections would be allowed
399. ons END waits for open connections to terminate on their own before terminating TQTCP No new connections are accepted after the END keyin has been received by TQTCP ABORT terminates TQTCP aborting any open connections Using ABORT may have unpredictable consequences during termination of the peer tqview or tqadmin end of the connection The END keyin is the preferred method for terminating TQTCP Examples TQTCP STATUS Prints a list of open connections on the system console TQTCP END Terminates TQTCP after open connections have closed TQTCP ABORT Terminates TQTCP after aborting any open connections TQ 01011 18 I 3 2200 Socket Interface 1 2 2 1 2 3 Processor Call The TQTCP program has the following processor call format TQTCP c config file elt k keyin name r u where C specifies the TQTCP configuration element By default the configuration element is located in SYS LIB TQ BASELINE TQTCP CONFIG At a minimum values must be provided in this configuration element for host user id password and account k specifies a keyin name for the TQTCP background run By default keyin name is TQTCP r runs TQTCP in real time mode Each connection made to the socket will operate in real time mode When using the r option both the account for the TQTCP background run and the account on the account statement in the TQTCP configuration element must allow real time mode u specifies to use u
400. optional probes to be controlled by tqdaemon Optional Probes tqsyb tqora i eg Eas EB End of Unix specific settings Locale specifies the language setting to be used at runtime It can be either C for English or a platform specific identifier for Japanese For 2200 Series systems Use ja to specify Japanese as the locale TeamQuest Baseline Administration For Unix systems Use the same value that is specified on the LANG environment variable setting for that platform eg ja japan or japanese Changing the setting to specify Japanese enables support for multibyte characters and causes messages to be output in Japanese Locale C Database For 2200 Series systems Data Directory specifies the qualifier to be used for all database files A dollar sign is appended to Data Directory to make the actual qualifier eg tqdata For Unix systems Data Directory specifies the directory where the database files should be stored A relative or absolute path name can be given Absolute path names must start with a Relative path names are relative to the parent directory of the executable Be sure there is no white space at the end of the line Data Directory tqdata Database specifies the default database to be used by all of the executables Database Production Aggregation Sets Aggsetl amp Aggset2 specify the aggregation sets that are to be directly populated by the probes There can be o
401. or example you feed a 1 month aggregation set from a day aggregation set that contains only 14 days worth of data each point in the month aggregation set will reflect an average for the last 14 days of the month not the average for the entire month Compact and Expanded Aggregation Set Format Aggregation set data is stored in disk files on the host system The aggregation set files can be in either compact or expanded format By default all aggregation sets created using TeamQuest Performance Software Administration or the tqasa program are in compact format The advantage of compact format is that significantly less mass storage is required to store the aggregation set data Data points are stored as single precision floating point values in compact aggregation sets Compact aggregation sets provide accurate data values while using minimal disk space to store data TQ 01011 18 6 7 Customizing Aggregation Sets 6 3 With compact aggregation sets no standard deviation information is saved with each data point You can only report the standard deviation when you are using time and or date consolidation of data points at report time When using the tqagg program to aggregate data from a compact aggregation set into an expanded aggregation set of a coarser granularity standard deviation information is inserted into the expanded aggregation set as each point is written An expanded aggregation set provides more complete reporting of stand
402. ormance applications such as TeamQuest View TeamQuest Alert TeamQuest On the Web and TeamQuest Model as needed This structure allows you to deliver the right performance information to the right people at the right time The TeamQuest Performance Software allows you to TQ 01011 18 Monitor large numbers of systems from a single console Investigate problems from the same tool with a few clicks of the mouse Use alarms to detect complex situations and rules of thumb for casual monitoring Typical users would include operators and system administrators Monitor one or more systems in detail with graphical and textual reports and use a powerful drill down facility to investigate problems Use correlation analysis to detect cause and effect situations Typical users include system administrators and performance analysts Analyze historical data with the powerful time selection and reporting capabilities including trend analysis Typical users are performance analysts and capacity planners Characterize workloads along business lines to understand and monitor resource usage and requirements Typical users are performance analysts and capacity planners Monitor systems from anywhere you have a standard web browser Automatically publish current and historical performance reports using your web browser Typical users are performance analysts remote or on site help desk and management Model your systems for hardware sizing capacit
403. ormat and layout of the report During a TeamQuest View session you can change the view type and layout of a report using the menu options and commands available through TeamQuest View This appendix provides a list of the predefined reports for your 2200 host system and a brief description of the contents of each report Report Definition Directory The report definitions are copied into a directory named report during installation The report directory is organized into a hierarchy of subdirectories The subdirectory structure is organized as follows report 2200 sip io m java memory proc system iotrace Summary mips mprlog syslog tpmlog wkldperf TQ 01011 18 B 1 Predefined Report Descriptions B 3 B 3 1 The report directory contains a subdirectory for the system in this case the 2200 The 2200 subdirectory contains a subdirectory for each available probe The sip subdirectory contains subdirectories for input output io memory processor proc system and summary reports The iotrace subdirectory contains IOTRACE reports The mips subdirectory contains MIPS probe reports The mprlog subdirectory contains MAPPER log reports The syslog subdirectory contains system log reports The tpmlog subdirectory contains TMP log reports The wkldperf subdirectory contains reports for workload service levels and facility utilizations Note If you are using TeamQues
404. oss the graph at the points of the horizontal axis designated by the minor ticks The acceptable values are ON or OFF If MinorGrid is set to ON the grid lines appear on the report If MinorGrid is set to OFF the grid lines do not appear on the report The default value is OFF The default values for XAxis Line are defined in the Axis Defaults settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences XAxis Enable Font Style Size Color The specifications for the text attributes of the labels for the horizontal axis The axis labels appear along the axis at the major tick marks and are used to identify the values of the horizontal axis Enable Exists in the syntax but does not affect the axis The functionality of this field has been replaced by the x axis labeling definition statement Font Specifies the typeface in which the labels are printed TQ 01011 18 Report Definition File Style Specifies the kind of emphasis for the typeface of the labels Some acceptable values are Normal For no emphasis Bold For highlighting the text Italic For emphasizing the text You can specify multiple styles for the labels However the styles must be joined by a plus sign For example to specify that the labels be printed in bold and italic typeface you enter bold italic in the Style field of the specification Size Defines the number of points for the typeface of the labels Th
405. ost that has the TEC SNMP Event Adapter tecad_snm or tecad_snmp running on it Also the TEC SNMP Event Adapter must be listening on port 162 which is the destination port for all CMS 1100 SNMP traps TQ 01011 18 J 7 2200 SNMP Trap Alarms 2 Configure the TEC SNMP Event Adapter to accept and process your alarm trap You can easily tailor the TEC SNMP Event Adapter to format a suitable Tivoli event from your trap and forward this event to an event server The event server then displays this event on the appropriate event console To configure the TEC SNMP Event Adapter you must update three configuration files as described in steps 3 through 5 3 Merge the following entries into the tecad_snmp oid file These entries should be added to the file so that the file remains in increasing object identifier order The following entries have been supplied to you in the symbolic element SYS LIB TQ BASELINE TECAD OID You can use this file to update the SNMP Event Adapter tecad_snmp oid file if TeamQuest 2200 MIB enterprises 132621 call df unisys 1s 3s6 lt 124 1 223 df os2200 1 3 62 1 451 223 7 tq2200 1 3 6 1 24 1 223 7 6 tq2200AlarmTrap 1 3 6 1 4 1 223 7 6 1 tq2200AlarmSystem i Sivbcd v4 Sl 223 Ly E tq2200Alarmidentifier 1336 Al 223 7 S 2 tq2200AlarmSeverity M3 Ors de Als LEAP 30s 3 tq2200AlarmValue Ow Ox dA LL 22367 7601 AP tq2200AlarmThreshZones Odi S E
406. ou have selected a table tqzip displays the following Retention Specification Menu 11 6 1 1 2 3 4 5 98 99 Enter Retention Specification Menu 2200 Runs List Retention Speci fications Modify Primary Retention Specification Add a Retention Speci fication Delete a Retention Specification Set Retention Start Omit Retention Speci Time fication Changes Save Changes and Exit Retention Specification Menu your selection lt 1 gt Listing Retention Specifications To display a list of the currently defined retention specifications for the table select List Retention Specifications from the Retention Specification Menu see 11 6 by entering 1 at the prompt The tqzip program displays the following list gt gt gt gt gt gt gt gt gt Ww gt gt gt gt gt Enter your selection lt I gt 1 Retention Start Time A 8 00 Retention Specifications 4 B Number Interval Keep C Primary raen 1 day D 1 00 01 00 1 week 2 00 30 00 1 month 3 01 00 00 6 months Continue The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A TQ 01011 1 The first line indicates the retention start time for the table This time is used for synchronizing data collection In this example the start time is 8 00 This indicates the number of retent
407. ou to create your own statistics that are derived using data values collected by the TeamQuest Baseline data collection probes You create derived statistics using TeamQuest View Once you have created a derived statistic it can be used in history monitor correlation and host based reporting You can also use the tqzip program to define thresholds for your derived statistics The thresholds are used by tqalm to determine when alarm messages need to be generated for a derived statistic A derived statistic is saved in the performance database as a formula The actual data values reported for a derived statistic are not stored in the database but are calculated at report time This has the following advantages e You can report on a statistic that is not being collected by the existing probes e You can report on a derived statistic for any time period that contains the data needed to calculate its value In a sense the derived statistic exists from the first day you collect data e You can modify a derived statistic formula and report on the same time period as before however this time you see data values reported that are based on the new formula e Because the data values are not stored minimal mass storage space is required for derived statistics No standard deviation is available for reporting with derived statistics Also no data is reported for a derived statistic if the formula could not be solved or if any of the components o
408. ouble The number of conversational sends issued in this sample interval N2PC_Calls Double The number of service calls issued to other services that did not use two phase commit in this sample interval N2PC_Regs Double The number of service requests processed that were not within a two phase commit in this sample interval 2PC_Calls Double The number of service calls issued to other services that used two phase commit in this sample interval 2PC_Reqs Double The number of service requests processed that were within a two phase commit in this sample interval Count Integer The number of combined occurrences This is collected when STATS_LEVEL gt 2 Txn_Commits Double The number of transactions committed that were initiated by this service in this sample interval Txn_Rollbacks Double The number of transactions rolled back that were initiated by this service in this sample interval Txn_Timeouts Double The number of transactions that timed out that were initiated by this service in this sample interval TQ 01011 18 A 65 Appendix B Predefined Report Descriptions B 1 B 2 Introduction Each of the probes released with the Team Quest Performance Software includes a set of predefined reports With Team Quest View you view these reports with data from the performance database The reports are defined in the report definition files The definition file determines the parameters displayed in the report and in addition the f
409. oughput Demand Runs Opened System Throughput Batch Runs Opened CINGI IG CD CD CG IG GG CD ED C2 ij e Re no E ae E Sa aD E oe S The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A TQ 01011 18 The first two lines of the file contain heading information The second line shows the parameter hierarchy of key names It is followed by a blank line Each parameter is listed by its key name hierarchy system category group category subcategory statistic resource application group workload set and workload All key values must be specified in each parameter reference An asterisk can be used as a mask indicator instead of specifying individual key values in the parameter references The asterisk can be used for any of the keys to simplify the deletion of sets of parameters For example all parameters that are for a given disk name can be deleted by specifying an asterisk in the statistic key and the disk name as the resource key along with all of the other required key values Customizing Performance Databases 10 9 Removing Database Parameters from a Database You can remove unwanted database parameters from a database For example if you have removed a particular disk device from your system you may no longer need the data collected for that device To remove database parameters usi
410. ould allow the tqbsp runstream to execute continuously as a background run on your system This ensures that your performance database contains a complete set of data for each day and that the real time monitor can access the most current data tqbsp Processor Call Statement The tqbsp runstream contains the tqbsp processor call statement which controls SIP data collection The tqbsp processor call statement specifies data collection parameters including input parameter table name database name and aggregation set names Normally you use the default values for these parameters as provided in the tqbsp runstream To modify any of the defaults you need to update the tqbsp runstream in SYS LIB RUN The tqbsp processor call has the following format tqbsp L r i parameter table d database name L a aggsetl L a aggset2 END where r executes tqbsp in real time mode The account for the tqbsp run must have a maximum real time level greater than zero i specifies the location of the input parameter table name to be used by the tqbsp probe For the tqbsp probe the parameter table is the element SYS LIB TQ BASELINE SIPINPUTTBL This is a required parameter on the tqbsp call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini element is used Make s
411. oup se workload set a workload oe resource et K END DERIVED STATISTICS L The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The derived statistics output file starts with a BEGIN DERIVED STATISTICS header line This line must not be changed or deleted B A blank line preceded by a pound sign must follow the header line TQ 01011 18 8 19 Customizing Derived Statistics C This is the first line of a derived statistic definition Each derived statistic definition consists of four sections Statistic Identifier Informational Text Formula and Aliases This line begins the statistic identifier section The statistic identifier specifies the name of the derived statistic and shows where it is stored in the database On subsequent lines following the section title are the key names for the derived statistic Each of the category group category subcategory and statistic key names must appear on a line by themselves and the actual values must be enclosed in double quotation marks D This line begins the informational text section Descriptive text about the derived statistic follows on subsequent lines The text must be enclosed in double quotation marks E This line begins the formula section The formula for the derived statistic follows on subsequent lines It must be enclosed in double quotation marks
412. oves the 30 minute aggregation set from the default database Any remaining aggregation sets are listed for your information gt tqasa r 30 minute gt TQASA 7R4G 011312 11 04 45 Fri Jan 13 14 59 08 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt Aggregation set 30 minute removed gt Aggregation Set List gt gt Max gt File gt Name Gran Start Stop Keep Size Source Name gt eaten ee Tiere ee ES Sikes a esheets heehee ee Sete ee tae gt 1l minute 00 01 00 00 00 00 23 59 59 08 00 00 23024 Trk gt gt gt gt An following Name indicates an expanded aggregation set gt gt End TQASA TQ 01011 18 6 19 Customizing Aggregation Sets 6 4 3 6 20 Listing Aggregation Sets The tqasa program allows you to produce a list of the aggregation sets currently defined for a database The tqasa program displays the aggregation set name granularity start time end time keep count maximum file size and the source name of the aggregation set To list the existing aggregation sets for a database the tqasa processor call has the following format tqasa d databasename where d specifies the database for which you want to list the existing aggregation sets The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 Example
413. ovided to assist in the maintenance and recovery of table files It may become necessary to reorganize your table files when any of the following occur e A table has a problem as indicated by an error termination from a probe or utility with one of the following error messages Error Invalid record sentinel detected Error Invalid record heading in file lt filename gt Error Invalid record type in file lt filename gt Error Internal error in index selection on table e Any problems with the table index such as a selection problem e A probe or utility appears to be performing poorly or the file size appears to be unusually large Every table file should be reorganized periodically if possible Performing the reorganization at least once a month is a good starting point The amount of data you collect and how frequently the tqagg program is run can affect the frequency of reorganizing the table data TQ 01011 18 11 12 1 Customizing Tables Why Do I Need To Reorganize My Table Files A number of factors enter into the need for periodic reorganization Because the actual reorganization can be a time consuming process it cannot be performed automatically on an on going basis A table file needs to be reorganized for the following reasons TQ 01011 18 As records are deleted from tables the space is left available for reuse as new records are inserted Fragmentation of the table space can occur over a period of time Reor
414. p tqrep f alarm rep end E 3 Report Definition File Format A report definition file is used to define the content and layout of the report generated by tqrep A report definition contains two or more sections a single Report section and one or more Table sections A section consists of the section name followed by section keys with assigned values The following conventions apply to a report definition file e The section names are enclosed in brackets e A section continues until the next section or the end of the report file is reached e The lines in a report definition can be continued to the next by using a backslash as the last character in the line e If strings in the report definition must be enclosed in double quotation marks a double quotation mark can be placed in the quoted string by placing a backslash in front of the embedded quotation mark For example filter login root TQ 01011 18 E 3 Host Based Reporting for Table Data The following example shows the report definition file alarm rep which was used to generate the report in Example E 1 Report A OutputFile alarm txt B StartTime amp NDATE 000000 EndTime amp NTDATE PageLength 20 PageWidth 80 CenterHeader Alarm Summary for amp BDATE CenterFooter amp PAGE Table C axRecs 500 D DataFields Timestamp MMDD hh mm Timestamp DataFields Severity LJ9
415. p Statistic Name Total IOs Sec Total IOs Sec MIN Total IOs Sec MAX Disk IOs Sec Tape IOs Sec Avg RET FLD Avg KWords Req Avg Xfer Time KWords Sec Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec KWords Req KWords Sec Transfer Time Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec Percent Reqs Read Percent Reqs Queued KWords Req KWords Req In KWords Req Out IO Activity Total I Os per second Minimum total I Os per second Maximum total I Os per second Total disk I Os per second Total tape I Os per second Average service time Number of K words per request Average transfer time Average K words per second IO Activity Physical Channel PCO PC1 Summary Physical channel percent busy Physical channel request per second Physical channel K words per request Physical channel K words per second Physical channel transfer time in milliseconds IO Activity Logical Channel LCO LC1 Summary Logical channel percent busy Logical channel request per second Logical channel percent requests that were reads Logical channel percent of requests that were queued Logical channel K words per request Logical channel K words per input request Logical channel K words per output request TQ 01011 18 OS 2200 Parameters Hit Rate Transfer Time RET FLD Hardware Serv Time Concurrency Category Group Category Re
416. p 3 tq2200AlarmValue OBJECT TYPE SYNTAX DisplayString SIZE 0 24 ACCESS read only STATUS mandatory DESCRIPTION The sampled value that caused the alarm to be raised This value is sent in display since we do not have a double type encoding for SNMP tq2200AlarmTrap 4 tq2200AlarmThreshZones OBJECT TYPE SYNTAX DisplayString SIZE 0 80 ACCESS read only STATUS mandatory DESCRIPTION This string depicts the theshold zones along with their boundaries icates normal and current The format of the field is lt valuel gt lt value2 gt Where represents Norm Curr or Prev Norm lt 45 00 gt Prev lt 75 00 gt Curr tq2200AlarmTrap 5 and ind Example tq2200AlarmDateTime OBJECT TYPE SYNTAX DisplayString SIZE 26 120 ACCESS read only STATUS mandatory DESCRIPTION The date and time that the alarm was generated in the format MM DD YYYY HH MM SS eg 01 13 2012 02 45 00 tq2200AlarmTrap 6 TQ 01011 18 2200 SNMP Trap Alarms tq2200AlarmTextl OBJECT TYPE SYNTAX DisplayString SIZE 0 64 ACCESS read only STATUS mandatory DESCRIPTION User text field 1 tq2200AlarmTrap 7 tq2200AlarmText2 OBJECT TYPE SYNTAX DisplayString SIZE 0 64 ACCESS read only STATUS mandatory DESCRIPTION User text field 2 tq2200AlarmTrap 8 tq2200AlarmText3 OBJECT TYPE SYNTAX DisplayString SIZE 0 64 ACCESS read only STATUS mandatory DESCRIPTION User text field 3
417. parameter a You can also set a relative time option when selecting the system log entries to process The format of the relative time option is t relative hour where t specifies a relative hour when selecting system log entries to process The valid range is 1 to 24 For example t 8 selects and processes the last 8 hours of system log data relative to when the tqsys program is started TQ 01011 18 Customizing Data Collection Probes 9 10 3 Workload Characteristics for the System Log Probe The system log probe has a unique set of characteristics that are used for its workload definitions You use these characteristics when you create the workload definitions using TeamQuest Performance Software Administration see 5 3 The following characteristics define the system log workloads Characteristic Name Run_ID Site_ID Program Transaction PID Account User_ID Project Trans_Code Lib_Bank Example 1 Valid Characters A Z 0 9 A Z 0 9 A Z 0 9 A Z 0 9 0 9 A Z 0 9 period A Z 0 9 period A Z 0 9 E A Z 0 9 gt 0 9 Valid Operations Type and Max Length string 6 SS string 8 string 12 string 12 integer 12 string 12 string 12 string 12 string 6 string 12 The following example shows a workload definition using run identifiers to identify system jobs This workload contains information for system background
418. parameters collected by any of the other TeamQuest Baseline probes the application group specification is null Data Values The term data value refers to a specific instance of data in the database for a specific parameter A data value is identified by the parameter aggregation set and time and date Based on this when you request values from the database you are requesting statistics for a particular time and date that were stored at a specified granularity Note Data for an interval t to t 1 is stored at time t Therefore when you see a data value at 10 00 a m it represents the activity for the interval starting at 10 00 a m In the case of snapshot values such as disk space available the data value at 10 00 a m represents the disk space available at the end of the interval started at 10 00 a m TQ 01011 18 2 3 TeamQuest Baseline and TeamQuest Online Concepts 2 2 4 2 3 2 4 Table Data The term table data refers to the data that is stored in tables A table is a collection of records containing columns that are related to one another in some way For example all data about an alarm is stored in a table A record contains one set of column values for the table For example data for a particular alarm on a system at a given time makes up one record in the alarm table Each field in a table contains a particular type of data The field may contain a string or a number The numbers may be integers real numbers elaps
419. pe Examples of numerical characteristics are PID system log probe and Appl_Group TPM log probe e Numerical operators can be lt gt lt and gt e A numerical value set is a single integer a real number or any numbers separated by commas and enclosed in braces Example 1 The following example will return values where PID is less than or equal to 123 PID lt 123 Example 2 The following example will return values where PID is equal to any of the values within the braces PID 1001 1012 1051 1022 TQ 01011 18 5 3 Customizing Workloads 5 2 2 2 Using String Conditions A string condition is expressed as a characteristic name a string operator and a string value set An example of a string condition is User_ID STF In this example User_ID is the characteristic name is the string operator and STF is the string value set Run_ID Site_ID and Program are examples of string characteristic names associated with the system log probe When creating a string condition the following rules apply e The characteristic name can be any of the characteristics associated with a specific probe listed as a string type e String operators can be and e A string value set can be any of the following A single string with the characters enclosed in double quotation marks or in single quotation marks Asingle pattern consisting of a regular expression enclosed in slashe
420. percent busy Control unit request per second Percentage of all requests that were reads Percentage of all requests that were queued Number of K words per request Number of K words per request that were input A 27 OS 2200 Parameters A 28 KWords Req Out Hit Rate Transfer Time RET FLD Hardware Serv Time Concurrency Response Time KWords Sec Avg Queue Length Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec Percent Reqs Read Percent Reqs Queued KWords Req KWords Req In KWords Req Out Hit Rate Transfer Time RET FLD Hardware Serv Time Concurrency Response Time KWords Sec Avg Queue Length Channel Queue Time Category Group Category Resource Subcategory Statistic Name Percent Busy Reqs Sec Number of K words per request that were output Hit rate Control unit transfer time Control unit service time Control unit hardware service time Control unit concurrency Unit average response time in milliseconds K words transferred per second Average number of requests queued to this unit IO Activity Disk DSKO DSK1 Summary Disk percent busy Number of requests per second Percentage of all requests that were reads Percentage of all requests that were queued Number of K words per request Number of K words per request that were input Number of K words per request that were output Hit rate Average transfer time for a request Average service
421. performance monitoring tool Scope This guide describes the TeamQuest Baseline and TeamQuest Online software It provides conceptual information for TeamQuest Baseline and the performance database structure It tells how to use the probes to collect and store performance data in the performance database Audience This guide is intended for those who are interested in performance analysis and capacity management of OS 2200 systems TeamQuest Online addresses the needs of the performance analyst TeamQuest Baseline addresses the needs of the performance analyst and the capacity planner The performance analyst who is typically interested in detailed data can use TeamQuest Baseline or TeamQuest Online to collect system data at a fine granularity The capacity planner can use TeamQuest Baseline to collect data across long periods of time Marketing representatives and technical managers who are interested in the performance management of OS 2200 systems might also use TeamQuest Baseline TQ 01011 18 jii About This Guide Prerequisites This guide assumes that you have basic knowledge of the Software Instrumentation Package SIP TPM log IOTRACE the MAPPER accounting log the system log and other existing performance analysis software on the OS 2200 system How to Use This Guide This guide is intended to be used with the TeamQuest View User Guide to view TeamQuest Baseline and TeamQuest Online reports The OS 2200 TeamQuest Baseli
422. ptions of tqzip see 8 4 1 e loads the new derived statistic definitions from the derived statistic definition file without first deleting the existing definitions in the database If you do not specify the e option all existing definitions are deleted from your database before the new ones are loaded TQ 01011 18 8 21 Customizing Derived Statistics Example The following shows an example of loading derived statistic definitions from the file dfile gt tqzip 1 s dfile gt TQZIP 7R4G 011312 11 04 45 Fri Jan 13 14 26 04 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt E database TQDATA PRODUCTION D loading derived statistics from lt dfile gt derived statistics were loaded Ta TQZIP 8 22 TQ 01011 18 Section 9 Customizing Data Collection Probes 9 1 9 2 Introduction TeamQuest Baseline and TeamQuest Online collect data from various sources on your OS 2200 system TeamQuest Online is released with a single probe for SIP data collection that executes as a separate batch run on your system TeamQuest Baseline is released with probes for SIP system activity monitor MIPS Metering ODTP TPM log IOTRACE file MAPPER log and system log data Each of these probes executes as a separate batch run on your system The TeamQuest Baseline and TeamQuest Online probes store data into the performance database The default database is found on the Database setting in the tq ini elem
423. pts per second RCS generic stack underflow overflow interrupts per second Signal interrupts per second Test and set interrupts per second Invalid instruction interrupts per second Page exception interrupts per second Arithmetic exception interrupts per second Data exception interrupts per second Operation trap interrupts per second Breakpoint interrupts per second Quantum timer interrupts per second Interrupt 21 not used interrupts per second Interrupt 22 not used interrupts per second Interrupt 23 not used interrupts per second TQ 01011 18 OS 2200 Parameters Software Break Software break interrupts per second Jump History Full Jump history full interrupts per second Delayed Hardware Delayed hardware check interrupts per second Check Dayclock Dayclock interrupts per second Performance Performance monitoring interrupts per second Monitoring Initial Program Load Initial program load interrupts per second UPI Initial UPI initial interrupts per second UPI Normal UPI normal TQ 01011 18 A 13 OS 2200 Parameters A 2 2 2 Dedicated Java Processor Statistics A 14 The following statistics are available on systems where it is possible to designate a processor or set of processors as dedicated Java IPs Parameter Hierarchy Category Group Category Subcategory Statistic Name Available Time Available Category Group Category Subcategory Statistic Name Idle Idle MIN Idle MAX Total Exe
424. r 1 at the prompt to register a table The tqzip program displays the following prompts gt Enter your selection lt 1 gt gt 1 gt gt Table filename i A gt savefile gt gt File gt gt 1 gt savefile contains tables B 2200 Runs gt Registering table 2200 Runs gt New table class C gt 2200 Archive gt New table name D gt Runs gt Table 2200 Archive Runs registered E The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A B TQ 01011 18 Enter the name of the file that contains the table you want to register The tqzip program prompts you with a list of tables included in this file Because the file in this example contains only one table it provides a message that it is registering the table 2200 Runs The tqzip program prompts you for a new table class The tqzip program prompts you for a new table name In this example the table called 2200 Runs in the file savefile is registered in the TeamQuest Baseline table list under the name 2200 Archive Runs When selecting data from this table for reporting you would choose the table 2200 Archive Runs The file name specified should be the complete name of the file When a table is registered the table name in the file is changed Thus if this file is registered again the list of tables in the file will be changed to reflect the previo
425. r administration tasks and terminated the open sessions close the TeamQuest Performance Software Administration window by selecting File gt Exit 4 8 TQ 01011 18 Section 5 Customizing Workloads 5 1 5 2 Introduction TeamQuest Performance Software Administration or the tqzip program can be used to administer the workloads in your performance database If you want to customize your workload definitions you can perform the following administrative tasks e Create workload sets and definitions e Modify workload definitions e Delete workload sets and definitions e Enable workload sets e Disable workload sets Concepts The system activity monitor TPM log probe real time TPM log probe IOTRACE MAPPER log and system log probes use the workload definitions for creating parameters and assigning data values Each probe has a unique set of workload characteristics The language to specify the workload definition is the same for all workload probes See 9 4 4 for the system activity monitor workload characteristics 9 7 5 for the RTPM and TPM workload characteristics 9 8 3 for IOTRACE workload characteristics 9 9 5 for MAPPER workload characteristics and 9 10 3 for system log workload characteristics The order that workloads are defined is very important When a probe processes an input record the workloads are evaluated in the order specified when they are defined Once a probe determines that the data record belongs to
426. r and from the same SIP session You can process SIP tape files that have been previously assigned to your run by specifying the file name used to assign the tape S specifies the system name that identifies the system source of the SIP data It is used to determine where to store the data in the performance database The system name parameter is case sensitive The tqbsp d and a parameters see 9 3 1 are also used for SIP file input Only one aggregation set a can be specified when using SIP file input and the a parameter is required on the tqbsp processor call If you do not want tqbsp to load the SIP data into the version of the performance database specified on the Database setting in the tq ini file use the d parameter to specify an alternate database version Note Before loading data into an alternate version of the performance database you must use the tqdbu utility to create this alternate version see 10 4 The tqbsp program processes each SIP file storing data points in the performance database as specified by the aggregation set parameter a If the SIP file terminates prematurely that is if the last block is not a SIP summary block the last data point is not written to the database If the last SIP block of a file is a SIP EOF and another SIP file is specified on the tqbsp processor call statement the next SIP file is processed The granularity of the input SIP file must not be greater than the granularity of the
427. r each day of the month First Day list box specifies the day of the week that should be considered the first day of the week This list box is only available when the Type is Week Exclude button specifies the days of the week to exclude from data consolidation This button is only available when the Type is Week or Month When you click the Exclude button the Days to Exclude dialog box is displayed Days to Exclude MV Sunday l Monday l Tuesday l Wednesday __Cancel_ l Thursday Friday MV Saturday When you check any of the boxes the host software excludes the data from those days when consolidating data 6 12 TQ 01011 18 Customizing Aggregation Sets Data Collection Times The Data Collection Times box contains a Range and an Interval scroll bar to make the time specifications for the data in the aggregation set Range scroll bar adjusts the time range of your data that is the start and end times for data collection Drag the scroll box or use the arrows on the scroll bar to specify the time range of your data The time range adjusts at the rate specified by the Interval scroll bar For example if the Interval is 4 hours the time range will be a period of 4 hours 08 00 to 12 00 09 00 to 13 00 and so on If the time range covers a period that crosses midnight an indicator appears next to the end time stating that the end time is on the next day Interval scroll bar allows you to adjust the time inter
428. r second Average dynamic bank size in K words Total dynamic bank loads per second The following statistics are available for Exec level SB6 and higher Parameter Hierarchy Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec TQ 01011 18 Memory Program Load TIP Trans Abs Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second Memory Program Load TIP Trans ZOOMS Average initial program size in K words Total initial program loads per second Average initial bank size in K words Total initial bank loads per second Average dynamic bank size in K words Total dynamic bank loads per second A 21 OS 2200 Parameters A 2 4 Paging Statistics A 22 The memory utilization statistics are not available from the Exec on OS 2200 on EM64T Server systems As the statistics will exist in the SIP data file with a value of 0 Team Quest Baseline will report zeros in the display of this data The following statistics are available for the 2200 systems on
429. r the aggregation set This value can be SUNDAY MONDAY TUESDAY WEDNESDAY THURSDAY FRIDAY or SATURDAY A value is allowed for this field only if the Group_Type is WEEK If no value is specified SUNDAY is used by default Exclude_Days The days of the week to exclude from data collection for the aggregation set This list of values can include any of the following SUNDAY MONDAY TUESDAY WEDNESDAY THURSDAY FRIDAY or SATURDAY The values are separated by a space or a comma For example to exclude Sunday and Saturday from data collection you would specify Exclude_Days SUNDAY SATURDAY A value is allowed for this field only if Group_Type is WEEK If no value is specified no days are excluded by default Compact Indicates whether or not this is a compact aggregation set If it is a compact aggregation set specify YES Otherwise specify NO If no value is specified YES is used by default Keep_Count The amount of data that is kept for the aggregation set It is expressed as n nMIN or nHOUR where n is the number of groups minutes or hours for which data is retained If only n is specified the length of a group is defined by Group_Type Do not put a space between the number and the unit A value for this field is required Fed_By The name of the aggregation set that feeds the aggregation set being created The specified aggregation set must already exist in the database You must use the Fed_By field if you want this aggrega
430. ram Interface product and make the related changes to your CMS 1100 or CPComm configuration Then start the Communications Application Program Interface background run on your OS 2200 system If you use an encrypted user id or password in the TQTCP configuration element you must have the Unisys Cipher Application Programming Interface API product installed on your system See I 2 3 and I 2 5 1 for additional information Update the SYS LIB TQ BASELINE TQTCP CONFIG element with the proper configuration information for your site Then start the TQTCP background run on your OS 2200 system Open a socket connection to your OS 2200 system from TeamQuest View or TeamQuest Performance Software Administration 2200 Socket Interface I 1 1 2 Communications Application Program Interface Communications Application Program Interface is a Unisys product that is bundled with releases HMP IX 4 0 and later It is an extended mode subsystem and a background run that provides an application interface to a communications network Access to the network is made through the TSAM interface of CMS 1100 or the API interface of CPComm using either TCP IP or UDP protocols Communications Application Program Interface gives an application program the ability to access the network by using simple commands such as LISTEN CONNECT SEND RECEIVE and DISCONNECT A background run performs the basic network control functionality through CMS 1100 and CPComm
431. rameter gt Enter your selection gt 1 gt gt Create threshold Update threshold or Exit c u lt e gt A gt u gt gt Current Threshold Definitions B gt 1 CLO1 10 Activity Disk I0s Sec gt 2 CLO1 10 Activity Disk Summary 10s Sec DSKF1 gt 3 CLO1 10 Activity Disk Summary Percent Busy DSKF1 gt gt Enter number name lt blank gt for more or E to exit C gt 3 7 12 TQ 01011 18 Customizing Alarms and Thresholds The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A To update a threshold enter u at the prompt B The tqzip program displays a list of the parameters that currently have threshold definitions c Select the parameter for which you want update the threshold definition You can enter either the number of the parameter or the full parameter name If the parameter list is longer than will fit on one screen enter a blank line to view more of the list After you select a parameter name the tqzip program displays the series of threshold definition prompts with the current value set as the default for each prompt see G through P in 7 4 You can modify any of the fields To use the default press the transmit key without entering a value at the prompt 7 7 Viewing Thresholds The tqzip program allows you to view an existing threshold definition Select View Thresholds from the Alarm
432. rcent Reqs Queued gt 209 Percent Reqs Read gt 210 Privileged Instruction gt 211 Program File Loads Sec gt 212 QPR active requests gt 213 QPR inputs queued gt 214 QPR max active requests gt 215 QPR spooled requests gt 216 QPR waiting requests gt gt Enter number name lt blank gt for more or E to exit gt 206 gt gt Full Statistic Name gt 1 10 Activity Control Unit Summary Percent Busy gt 2 10 Activity Disk Summary Percent Busy gt 3 IO Activity Logical Channel Summary Percent Busy gt 4 10 Activity Other Summary Percent Busy gt 5 I0 Activity Tape Summary Percent Busy gt TQ 01011 18 Customizing Alarms and Thresholds gt Enter number name lt blank gt for more or E to exit gt 2 gt gt Resource Name E gt 1 DSKAO gt 2 DS gt 3 DS gt 4 DS gt 5 DS gt 6 DS gt 7 DS gt 8 DS gt 9 DS 0 0 NnOoOVUMVNMVWOS gt S WZ ANA nTFWMH NANNNNNNNN m O Ea a O oe oo oe G Enter number name lt blank gt for more or E to exit 0 SEU M SEME SE NENE NONSE NY gt gt CLO1 10 Activity Disk Summary Percent Busy DSKF1 F gt Enable threshold checking y lt n gt G gt y gt Number of threshold zones lt 2 gt H gt 3 gt Value is usually in zone lt 1 gt I gt 1 gt First threshold value lt 00000 00 gt J gt 50 gt Second threshold value lt 00000 00 gt K gt
433. rd TQ 01011 18 Customizing Data Collection Probes 9 10 2 System Log File Input With the tqsys program you can load data from a system log file into a performance database by specifying the date range or the specific system log file This enables you to load system log data from another system or to update the performance database with system log data that was collected at a time when tqsys was not running on your system When loading data using a date range tqsys searches the system and processes all system log files that contain information in the date range You must specify two additional parameters on the tqsys processor call statement See 9 3 1 for the format of the processor call statement The format for the date range parameters is the following b mmddyylyy e mmddyyl yy where b specifies the month day and year of the beginning date used in the log file selection e specifies the month day and year of the ending date used in the log file selection If a 2 digit year is specified the digits 00 through 27 are interpreted as the years 2000 through 2027 and the digits 28 through 99 are interpreted as the years 1928 through 1999 To load data directly from specific system log files you must specify two additional parameters on the tqsys processor call statement See 9 3 1 for the format of the processor call statement The format for these parameters is as follows f systemlog n L f systemlog ntl
434. re are a number of rules to consider when determining if an aggregation set can or should be fed by another aggregation set e The granularity of the fed aggregation set must be greater than or equal to the granularity of the source aggregation set A 15 minute aggregation set can feed a 60 minute aggregation set but a 60 minute aggregation set cannot feed a 15 minute aggregation set e The granularity of the fed aggregation set must be an even multiple of the granularity of the source aggregation set For example a 5 minute aggregation set can feed a 30 minute aggregation 30 minutes is an even multiple of 5 minutes e The date consolidation type of the fed aggregation set must be larger than or the same as the feeding aggregation set For example a week consolidation type can feed another week consolidation type but it cannot feed a day consolidation type e A week consolidation type cannot feed a month consolidation type as most months do not have an even number of weeks in them e The days to exclude must match between the fed aggregation set and the source aggregation set e Ifa week consolidation type feeds another week consolidation type the start day of the week must be the same for both aggregation sets e Ifthe source aggregation set does not retain data for a period that is long enough to encompass a single point in the fed aggregation set the fed aggregation set data will not accurately reflect the specified interval If f
435. rence would contain ALL_RES for the resource key specification This would expand into one reference for each disk on your system The AVG function would average all of these values and return a single value for the derived statistic Unlike individual substitution only one value is reported as a result of resolving a formula using group substitution In a single formula all parameter references that use substitution must be of the same type For example you cannot use RES in two parameter references unless they are both for the same type of resource for example either tape or disk This means that using both tape and disk substitution in one formula is not allowed Parameters using group substitution have no associated type and their usage is not limited as with individual substitution Formula Definition A derived statistic formula is used to calculate the value that is reported for the statistic This formula is entered when creating a derived statistic definition using TeamQuest Performance Software Administration Formulas can consist of parameter references by using aliases arithmetic operators constants and mathematical functions The MIN MAX AVG SUM and GRAN functions are supported The MIN MAX AVG and SUM functions require a list of values as input The value list is made up of one or more values separated by commas A value can be either a numeric constant or a parameter reference alias TeamQuest Baseline obtains the
436. required TQ 01011 18 9 63 Customizing Data Collection Probes Table Removal The syntax for executing tqtblprb to remove a table is filename tqtblprb r d databasename t class name end where r specifies a table removal execution d specifies the database the table is stored The databasename parameter is the name of an existing database on your system If the databasename is not specified the database in the tq ini file is used t specifies the table class and table name for removal This field is required Caution You should take special care when removing a table Once a table is removed all the data stored in the table is gone If the table was not backed up the data stored in the table can never be recovered The following example shows a table removal execution that removes the 2200 accountdisk table sys lib tq baseline tqtblprb r t 2200 accountdisk end 9 64 TQ 01011 18 Customizing Data Collection Probes 9 11 2 5 Keyword Definitions Time Types Definitions The time type in a TABLE definition see Table Definition in 9 11 2 1 can be one of the following Table_Time_Granulized is used for cumulative data Table_Time Sampled is used for sampled data This is the default Table_Time_Event is used for event data Field Types Definitions The field type in a FIELD definition see Field Definition in 9 11 2 1 can be one of the following Field_Type_None is th
437. resented are Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec and Deadlocks Sec MHUDSREQ RPT Multihost UDS Reqs Sec Provides total UDS requests per second information on Multihost usage Parameters presented are Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec and Deadlocks Sec MH TOT RPT Multihost Reqs Provides system wide total requests information on Multihost usage Parameters presented are Lock Reqs Unlock Reqs Delayed Resp Msg and Deadlocks MHTIPTOT RPT Multihost TIP Reqs Provides total TIP requests information on Multihost usage Parameters presented are Lock Reqs Unlock Reqs Delayed Resp Msg and Deadlocks MHUDSTOT RPT Multihost UDS Reqs Provides total UDS requests information on Multihost usage Parameters presented are Lock Reqs Unlock Reqs Delayed Resp Msg and Deadlocks LOCKTOT RPT UDS Lock Counts Provides total counts on the various UDS RLP lock functions These parameters are collected for XPCs when the UDSTRP SIP counter exists LOCKREQ RPT UDS Lock Reqs Sec Provides total counts per second for the various UDS RLP lock functions These parameters are collected for XPCs when the UDSTRP SIP counter exists Predefined Report Descriptions File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File
438. riod Menu see 11 5 indicates if data collection is to be done for the table If data collection is currently turned off the menu selection allows you to turn it on by entering 4 at the prompt If data collection is turned on the menu selection allows you to turn it off by entering 4 at the prompt By using this selection you can stop collecting data for a table while retaining the collection periods currently defined Removing all collection periods for a table would have the same effect as turning off data collection Exiting the Collection Period Menu When you exit the Collection Period Menu you can apply any collection period changes you made by selecting Exit Collection Period Menu Enter 99 at the prompt If you do not want to apply the collection period changes you can select Omit Collection Period Changes from the menu Enter 98 at the prompt gt Collection Period Menu 2200 Runs gt gt 1 List Collection Periods gt 2 Add a Collection Period gt 3 Delete a Collection Period gt 4 Turn Data Collection Off gt gt 98 Omit Collection Period Changes gt 99 Save Changes and Exit Collection Period Menu gt gt Enter your selection lt 1 gt gt 99 When you select 98 or 99 at the prompt the tqzip program returns you to the Tables Menu see 11 3 Defining Retention Specifications Retention specifications define how long data is kept for tables The first retention specification is for dat
439. ription 6 27 tqalm program alarm messages 7 30 alarms handling 7 28 command line format 7 28 description 7 28 processor call statement 7 28 user defined alarm handling 7 31 viewing alarm messages 7 27 tqasa program command line format 6 16 6 19 6 20 6 26 converting expanded aggregation sets 6 26 creating an aggregation set 6 16 listing aggregation sets 6 20 removing an aggregation set 6 19 tqbsp program abnormal termination 9 3 description 1 8 9 2 parameters collected by A 2 A 51 processor call statement 9 2 SIP file input 9 4 tqetl runstream 9 1 tqdata description 1 11 tqdbu utility creating databases 10 3 listing databases 10 4 TQENC program for encryption I 10 tqgetm program F 1 tqiot program description 1 9 9 25 parameters collected by A 44 processor call statement 9 26 workload characteristics 9 28 tqmips program abnormal termination 9 11 description 9 9 Index 7 Index parameters collected by A 43 processor call statement 9 9 terminating data collection 9 11 trail selection 9 11 tqmpr program changing default log filenames 9 32 description 1 9 9 29 log fileinput 9 31 MAPPER log file input 9 31 parameters collected by A 45 A 53 processor call statement 9 29 workload characteristics 9 33 tqodtp program abnormal termination 9 15 description 1 10 9 14 parameters collected by A 61 processor call statement 9 14 table data 9 16 terminating data collection 9 15 tqrep pr
440. rkload set to the host software you must click OK in the Workload Maintenance dialog box Your definition is not sent to the host software until you click OK Displaying or Modifying a Workload Set Definition To display or modify a workload set definition select the workload set from the Workload Info list box in the Workload Maintenance dialog box see 5 3 1 and click Open You can also double click on the workload set name in the Workload Info list box The Workload Set Definition dialog box is displayed 4 Workload Set Definition Session cl01 Database PRODUCTION Source Data Type hyslog Workload Set Name e IV Active Cancel Help You can use this dialog box to modify the workload set name active status and qualifiers For a description of the fields in this dialog box see 5 3 2 When you click OK your changes are accepted and you return to the Workload Maintenance dialog box see 5 3 1 Your changes are not sent to the host software until you click OK in the Workload Maintenance dialog box TQ 01011 18 5 11 Customizing Workloads 5 3 4 5 12 Deleting a Workload Set To delete a workload set from a database you select the workload set name from the Workload Info list box in the Workload Maintenance dialog box see 5 3 1 When you click the Delete button an alert box verifies your request to delete the workload set When you click Yes in this alert box the workload set name is removed from the Workloa
441. rmation about tqctl 7 28 TQ 01011 18 Customizing Alarms and Thresholds Processor Call Format The processor call format for tqalm is tqalm a aggset d database c L w delaysecs m size r where a specifies the aggregation set used by tqalm to obtain the data values needed to calculate the values for derived statistic threshold checking If the a option is not specified the AGGSET1 setting in the tq ini element see 3 1 is used The granularity of the aggregation set is used to determine how often to check the derived statistic thresholds d specifies the database that tqalm uses The database parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 The tqalm program uses this database to retrieve information for the derived statistic thresholds it needs to be checking C sends a console message for each alarm that is generated W specifies the delay factor in seconds to use in derived statistic threshold checking By default tqalm uses the most recent data point stored to calculate derived statistic values for threshold checking Using the delaysecs parameter gives all the TeamQuest probes time to write their data values before the derived statistic values are calculated for threshold checking The number of seconds specified as delaysecs is rounded up to the nearest whole number of samp
442. robe run has the best opportunity to report data as it is being written to the TPM audit trail The default is to not run in real time u processes the user area of TPM termination records using the LINC format described in 9 7 6 Z is used to register the tqrtpm probe in the database i specifies the location of the input parameter table name to be used by the tqrtpm probe For the tqrtpm probe the parameter table is the element SYS LIB TQ BASELINE TPMINPUTTBL This is a required parameter on the tqrtpm call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name TQ 01011 18 a Note Customizing Data Collection Probes specifies the aggregation set to be used for data collection The aggset1 parameter is the name of an aggregation set defined in the database name The aggset1 parameter is case sensitive and only one aggregation set is allowed The aggregation set is optional on the call line If an aggregation set is not specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used See 2 2 1 for a description of aggregation sets specifies the workload set names to be used in data collection If a workload set name is not supp
443. rs being collected by the data collection probes They can also be used to reference other derived statistics When an alias refers to a parameter being collected by the data collection probes the value for the parameter is retrieved from the database and is inserted into the formula at evaluation time When an alias refers to another derived statistic the value for the second derived statistic is calculated and the result is inserted into the formula for the first derived statistic at evaluation time Because you can use aliases to refer to other derived statistics in your formula be careful that you do not specify a circular definition that is a formula that either refers to itself directly or indirectly through the formula of a referenced derived statistic If this condition is encountered a warning dialog box informs you of the situation You must correct the problem before the derived statistic definition can be saved in the database In general the formula is evaluated from left to right The following operators are supported in order of precedence Parenthesis can be used to provide correct precedence in the formula Exponentiation aay Multiplication and division Addition and subtraction Numeric constants can be specified either with or without a decimal point However they are all treated as floating point numbers No integer operations are supported Both positive and negative constants are allowed Exponential notation
444. rther qualified by a resource specification A resource is a component of the system that does work A resource could be a physical resource such as a disk unit or tape unit Or it could be a logical resource such as a system call For example a statistic with a hierarchy of CL01 IO Activity Logical Channel Cache Summary Transfer Time can be further qualified by the specific logical channel of LC1 In this case TeamQuest Baseline or TeamQuest Online would report the transfer time statistics for LC1 of your system Workload Specification With the workload capability of TeamQuest Baseline a parameter can be further qualified by a workload specification A workload is a logical classification of the work performed on the system These classifications are site specific and are expressed in terms that are meaningful to the business For example workloads can be defined to track how much of the system resources are used by the payroll department the order entry department or a software development project Note The workload capability is not available with TeamQuest Online Application Group Specification A parameter can further be qualified by an application group specification For the MAPPER probe tqmpr the MAPPER name is specified using the application group specification For the UDS and MCB probes supplied with the TeamQuest Probes product the UDS or MCB application name is specified using the application group specification For
445. s Several strings and patterns separated by commas and enclosed in braces Example 1 The following example will match all occurrences where the string begins with TQ The part of the string is significant where the period is required with the asterisk to wild card match anything after TQ Run_ID TQ Example 2 The following example illustrates the matching of multiple strings separated by commas and enclosed in braces Run_ID HALE TSTERLO 9 SAA BDK In the preceding example any Run_ID satisfying any of the following strings or patterns will be returned HALE matches only on the string HALE TSTERLO 9 matches on the string TSTER and a single numeric number 0 through 9 SAA matches only on the string SAA BDK matches only on the string BDK TQ 01011 18 Customizing Workloads Example 3 A string condition containing a string enclosed in double quotation marks or single quotation marks indicates that an exact match is required when the definition is evaluated The following examples are interchangeable Program FURPUR matches on the string FURPUR and only FURPUR Program FURPUR matches on the string FURPUR and only FURPUR 5 2 2 3 Using Regular Expressions in String Conditions A string condition containing a string enclosed in slashes indicates the use of regular expression matching The string contained between the slashes can be
446. s 9 5 3 Abnormal Termination of Data Collection _ 9 5 4 Trail Selection By Date Range or Relative Time 9 5 5 System Log File Input ecceeeeeeeeeeeeeeesteeeeeees 9 6 Open Distributed Transaction Processing Collection 9 6 1 tqodtp Processor Call Statement eee 9 6 2 Terminating Data Collection cceeeeeeeeeeeeenees 9 6 3 Abnormal Termination of Data Collection _ 9 6 4 tqodtp Table Data occ eeecceeeeeeeneeeeeeeeneeeeeeeaae 9 6 4 1 System Table Information ccceeeeeeeeees 9 6 4 2 Server Table Information cccceeeeeeeeeee 9 6 4 3 Client Table Information ccccccceeeeceeeee 9 6 4 4 Service Table Information ccsseeeeeees 9 7 Real Time TPM and TPM Log Data Collection _ 9 7 1 tqrtpm Processor Call Statement 9 7 2 Terminating tqrtpm Data Collection 9 7 3 tqtpm Processor Call Statement eeeeees 9 7 4 TPM Log File Input ooo ee eeeeeeeeeeeeneeeeeeeenaes 9 7 5 Workload Characteristics for the RTPM and TPM Log File Probes aeeiiaii iaa 9 7 6 Log File User Area ssessrissssrrrssserirnnsrrirnessenns 9 8 IOTRACE Data Collection o eecccccecsseeceeesssteeeeeeeeeaes 9 8 1 Data Collection RUNStreaM ceeeeeeeeesteeeeeeeees 9 8 2 tqiot Processor Call Statement c cece 9 8 3 Workload Characteristics for the IOTRACE Probe 9 9 MAPPER Log Data Collection eseese 9 9 1 Data Collection Runstream ceeceeee
447. s J 9 J 7 Special Considerations eeceeeeeeeeeeseeeeeeeneeeeeteeeeeeneeees J 12 GIOSSANY E E A A E AT 1 Bibliography siainen eiee aaie e a e aa A a EE E A Ea 1 Sae ee e E E E T E E T 1 TQ 01011 18 xix Figures 1 1 TeamQuest Performance Software Family o ccsceeeccecseeeeeeeteeeseeeeeees 1 2 TeamQuest Baseline Components 6 1 Parameter Grouping for an Aggregation Set eececeeeeeeeeeeeeeeetteeeeneeees 6 2 Data Consolidation from Aggregation Set 1 to Aggregation Set 2 _ 7 1 Threshold Zones scets civeniccesiiidiendiel ies cned uate N teed E 1 Report Generated by tqrep ou eeeeeeeceeeeeeneeeceneeeeeaeeceeeeeeeaaeeseseeeeesaaeetenes TQ 01011 18 xxi Part 1 Getting Up and Running Section 1 Introduction 1 1 What ls TeamQuest Performance Software TeamQuest Performance Software is a family of products which addresses the needs of performance management and performance analysis of your computer systems This integrated family of products allows you to choose the tools that you need to manage the performance at your site The TeamQuest Performance Software family consists of the following products e TeamQuest Baseline TeamQuest SMFII and TeamQuest Manager The TeamQuest Performance Software includes TeamQuest Baseline for Unisys ClearPath OS 2200 systems TeamQuest SMFII for Unisys MCP systems TeamQuest Manager for UNIX Linux systems and TeamQuest Manager for Microsoft Windows systems The
448. s and Thresholds Menu see 7 3 by entering 2 at the prompt The tqzip program then prompts you with a list of parameter names that currently have threshold definitions in the database as shown in the following example gt Enter your selection gt 2 gt gt Current Threshold Definitions A gt 1 CLO1 10 Activity Disk I0s Sec gt 2 CLO1 10 Activity Disk Summary 10s Sec DSKFl gt 3 CLO1 10 Activity Disk Summary Percent Busy DSKF1 gt gt Enter number name lt blank gt for more or E to exit B gt 3 gt gt CLO1 10 Activity Disk Summary Percent Busy DSKF1 C Zif Threshold checking ON gt 2 Threshold 1 00050 00 Threshold 2 00075 00 po umber of zones 3 ormal zone 3 al gt 4 Secs before alarm 0060 Secs delay 0060 gt 5 Alarm identifier Percent Busy gt 6 User text When more than 50 busy call John at ext 2332 gt 7 User text gt 8 User text gt gt gt Do you want to view another threshold y lt n gt D gt TQ 01011 18 7 13 Customizing Alarms and Thresholds 7 8 7 14 The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The tqzip program displays the list of threshold definitions currently defined in the performance database Select the parameter you want to view You can enter either the number of the parameter or the full parame
449. s and their components as an add on to SPO For TeamQuest Baseline alarm messages to drive status displays the message and pattern matching considerations described previously apply Single Point Status uses Attribute Change AC and Delete DE event reports rather than Alarm AL event reports but it provides the capability to automatically create alarm event reports when a monitored attribute attains a user specified value or enters a user specified range In contrast with SPO Status diagrams show not only normal green and alarm red states but also an attention state yellow that can indicate a component whose performance is worse than normal but not yet of alarm significance If you use Status to show the SPO operator critical data reported by TeamQuest Baseline then you will probably find it best to use SMART SP AMS to create AC and DE event reports and let Status automatically generate the AL event reports See the SPO documentation for details of patterns pattern matching and actions event reports alarm help text and external alarm actions For more information on SMART SP AMS see the Single Point Autoaction Message System Administration Guide and the SMART for ClearPath OS 2200 Administration and Operations Guide Other Considerations You will need to define TeamQuest Baseline thresholds locally on each monitored system No centralized definition point exists However you may be able to use the threshold load unload capab
450. s available text The name of a TeamQuest Performance Software product to which the report definition applies Report Name text The name of the report as it appears on the title bar of the report window text Can be a string of characters up to 50 characters long If it contains spaces it must be enclosed in quotation marks Host Type host type The type of computer hardware with which this report can be used Each hardware type has a unique set of host defined key names The combination of keys is unique for a particular class of hardware This specification must equal the host type provided by the host product If this specification is made a check is done with the host as identified by the host product If a mismatch occurs the user can ignore the error or cancel Date Format format The specifications for the format of the date as it appears on the report format Can be any of the following and must be entered in uppercase YY Years 00 99 YYYY Years specified with the century MM Month 01 12 MMM Month abbreviated name MMMM Month complete name DD Day of the month 01 31 DDD Day of the year 001 366 DDDD Day of the week 0 6 where 0 Sunday WW Week of the year 0 53 TQ 01011 18 Report Definition File WWW Day of the week abbreviated name WWWW Day of the week complete name The default values for Date are defined in the Date Time Formats settings of the User Preferences See the
451. s command from the File menu to save a copy of this document TQ 01011 18 ix About This Guide To access product documentation from your product CD or DVD 1 Load the CD or DVD in your disk drive The disk contents dialog box is displayed Open the welcome pdf file The welcome pdf file can be used to navigate to the desired PDF file You may wish to copy the welcome pdf file and the pdfdocs folder to your server so multiple users have convenient access to the product documentation Note When copying the welcome file and the pdfdocs directory you must maintain the established directory structure If the directory structure is not maintained the navigation links within the welcome pdf file will not open the documents located in the pdfdocs directory Notation Conventions The examples in this guide show as nearly as possible how your input system output and screens appear on a terminal Program names are shown in boldface type In this guide the following conventions apply for command formats when using TeamQuest Baseline and TeamQuest Online on the host system You can enter any commands or other required input in either uppercase or lowercase letters Both ways are shown in the examples in this guide In command and format examples anything that appears in regular type must be entered exactly as shown For example in the following command tqbsp r i parameter table you must enter tqbsp exactly as
452. s information is only available if you have the hardware monitor option UTIL SUM RPT Java Processor Utilization Provides Java processor utilization statistics for user categories IDLE RPT Java Processor Utilization Idle Provides Java processor utilization statistics for idle as well as the minimum and maximum idle for the time range UTIL TIP RPT Java Processor Utilization TIP Provides TIP Java processor utilization as well as the minimum and maximum values for the time range Memory Reports The memory report definitions are found in the memory subdirectory of the sip subdirectory File Name Description File Name Description File Name Description File Name Description TQ 01011 18 AVL MEM RPT Available Memory Gives a summary of available memory LD INITB RPT Initial Bank Load Rate Provides initial bank load rates broken down by the various executable categories LD INITP RPT Init Program Load Rate Provides initial program load rates broken down by the various executable categories LD DYNB RPT Dynamic Bank Load Rate Provides dynamic bank load rates broken down by the various executable categories Predefined Report Descriptions B 3 4 File Name Description File Name Description File Name Description File Name Description PAGE WS RPT Paging Working Sets Provides paging working set information PAG
453. s processed per second Active seconds per display function start Display function starts per second Average number of signed on users Mapper Log Workload Total WLS1 WLS2 all I Os processed per second Logic and data lines processed per second Active seconds per display function start Display function starts per second Average number of signed on users A 53 OS 2200 Parameters A 8 3 System Log Probe Derived Statistics A 54 The following statistics are collected by the system log tqsys probe Parameter Hierarchy Category Group Category Subcategory Workload Set Workload Statistic Name Kmem txn XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec txn Throughput txns sec Population active txns Category Group Category Subcategory Workload Set Workload Statistic Name Kmem txn XPA Tape IOs sec Disk IOs sec IO SUPs sec ERCC SUPs sec IP SUPs sec Response sec txn Throughput txns sec Population active txns System Log Overall TIP WLS1 WLS K words of memory per TIP transaction for XPA systems TIP tape I Os per second TIP disk I Os per second TIP I O SUPs per second TIP ER CC SUPs per second TIP IP SUPs per second Seconds per TIP transaction TIP transactions per second Average number of active transactions System Log Workload TIP WLS1 WLS2 all K words of memory per TIP transa
454. s to any management console that accepts Simple Network Management Protocol SNMP traps vi TQ 01011 18 About This Guide Results After reading and using this guide you will be able to use TeamQuest Baseline or TeamQuest Online on the host system maintain the performance database and use the data collection probes to gather measurement data in the database Related Product Information The following related documents may be helpful to you when using TeamQuest Baseline and TeamQuest Online Refer to the version that corresponds to the level of software in use at your site TeamQuest Product Documentation If you are using TeamQuest manuals you can find related product information in the following documents TeamQuest Model User Guide for Unisys Systems TQ 01424 This guide describes how to use the TeamQuest Model product with Unisys systems TeamQuest Model allows you to build and solve models to predict the performance of your computer system should conditions change OS 2200 TeamQuest PMLog End Use Reference Manual TQ 01351 This manual provides information for using the TeamQuest PMLog software with TeamQuest Baseline OS 2200 TeamQuest Probes End Use Reference Manual TQ 01361 This manual provides information for using the TeamQuest Probes software with TeamQuest Baseline or TeamQuest Online OS 2200 TeamQuest Products Installation Guide TQ 01013 This guide contains information you will need to install any of the
455. sage Total number of I O requests made by this run Total number of I O requests to disk that have been made by this run Total number of words that have been transferred during disk I O operations by this run Total number of I O requests to tape that have been made by this run Total number of words that have been transferred during tape I O operations by this run Total voluntary wait time Sum of user response time and TWAIT time Total amount of time the run spent waiting for a resource Total time the job was active TQ 01011 18 OS 2200 Parameters Transactions Cards In Runs Opened Max Real Memory Real Memory Usage Category Group Category Subcategory Statistic Name Total SUPs IP SUPs ER CC SUPs ES IO SUPs ES IP SUPs CS IO SUPs CS IP SUPs IO SUPs Core Block SUPs IO Requests Disk I O Requests Disk I O Words Tape I O Requests Tape I O Words Voluntary Wait Resource Wait Active Time Transactions Max Real Memory Real Memory Usage TQ 01011 18 Total number of transactions Total number of cards the Exec read during the run Total number of runs that were opened Maximum number of real pages concurrently resident for this run The product of the total number of resident pages and the time in seconds that the pages are resident System Log Workload TIP Total SUPs used to process the job Sum of the IP SUPs ER CC SUPs operating system SUPs and I O SUPs Total IP S
456. se products provide the services required for all of the products in the TeamQuest Performance Software family These products provides the following features performance data measurement collection and storage services administrative interfaces alarm handling capabilities basic textual data access and reporting capabilities e TeamQuest Online This product provides a subset of the services provided by TeamQuest Baseline All other products in the TeamQuest Performance Software family require the TeamQuest Performance Software e TeamQuest Alert This product provides a performance console that allows you to monitor the current health of a large number of systems TeamQuest Alert obtains performance data and alarm events from TeamQuest Manager on monitored system It provides a top down approach to monitoring large numbers of systems with a tightly coupled drill down capability for problem investigation e TeamQuest Model TeamQuest Model is a client application that runs on your workstation TeamQuest Model provides capacity planning and what if analysis TeamQuest Model obtains its data from TeamQuest products running on Unisys ClearPath OS 2200 systems Unisys MCP systems and UNIX Linux systems It provides an easy to use interface to a set of powerful modeling algorithms TQ 01011 18 1 1 Introduction e TeamQuest On the Web This product provides performance reports using a browser TeamQuest O
457. section name is followed by section keys with their assigned values This line begins the Table section keys See E 3 2 for a description of the valid keys for the Table section TQ 01011 18 Host Based Reporting for Table Data E 3 1 TQ 01011 18 Report Section Keys The keys listed in this section are valid in the Report section of a report definition file The key names are not case sensitive Unless otherwise noted the key names can be specified only once If a key is specified more than once the last key value specified is Database database The name of the database to use The database must be enclosed in double quotation marks This specification can be overridden by the d option on the tqrep command line If the Database key is not specified in the report definition file and the d option is not given on the command line the default database is used OutputFile filename The name of the file where the report is to be written The filename must be enclosed in double quotation marks This specification can be overridden by the o option on the tqrep command line If OutputFile key is not specified in the report definition file and the o option is not given on the command line the report is written to the standard output file StartTime MMDDYYYY 4HMMSS The starting date and time for the report This specification can be overridden by the s option on the tqrep command line If the time HHMMSS
458. sename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 u specifies that an aggregation set unload operation is to be performed by tqasa All aggregation set definitions in the database are exported into the file specified by the aggset file parameter Example The following shows an example of unloading aggregation set definitions from the default database into the file aggsetfile gt tqasa u aggsetfile gt TQASA 7R4G 011312 11 17 30 Fri Jan 13 08 17 27 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserved gt gt unloading aggregation set definitions to aggsetfile gt 2 definitions were unloaded gt gt gt End TQASA TQ 01011 18 6 21 Customizing Aggregation Sets 6 22 Aggregation Set Definition File Format The aggregation set definitions appear in the following format in the output file created by tqasa If you choose to edit this file you should maintain this file format BEGIN AGGSETS unloaded on Fri Jan 13 08 17 28 2012 A Name 1 minute Granularity 60 Start_Time 000000 Stop_Time 235959 Group_Type DAY Week_Start SUNDAY Exclude_Days Compact YES Keep_Count 8HOUR Fed_By Name 10 minute Granularity 600 Start_Time 000000 Stop_Time 235959 Group_Type DAY Week_Start SUNDAY Exclude_Days Compact YES Keep_Count
459. ses An initial database is created at installation If you are content with the initial database no additional administration is required However if you want to customize your databases you can perform the following administrative tasks e Create a database see 10 4 e Delete a database see 10 5 e List databases see 10 6 e Estimating disk space requirements for a database see 10 7 e Generate a list of database parameters see 10 8 e Delete parameters for a database see 10 9 e Clearing a locked database see 10 10 e Backup a database see 10 11 e Recover a database see 10 11 10 2 Performance Databases TeamQuest Baseline and TeamQuest Online use the performance database to store control information as well as the system performance data collected on your system Before you can use TeamQuest Baseline or TeamQuest Online the performance database must exist on your system The initial database is created automatically when you install TeamQuest Baseline or TeamQuest Online The initial performance database is named PRODUCTION After this initial database is created you can create additional databases to fit your particular needs You can also delete any databases when they are no longer needed TeamQuest Baseline and TeamQuest Online contain a utility program called tqdbu that handles the creation and deletion of performance databases TQ 01011 18 10 1 Customizing Performance Databases 10 3 Database Files
460. set and the number of workload set qualifiers is 1 c The second line of the workload set definition contains the name of the first workload set qualifier and the number of qualifier values for the workload set qualifier The qualifier is System and there are three qualifier values D Each qualifier value for the workload set is then listed on a line by itself Venus saturn and mars are the qualifier values for this workload set E The second workload set definition begins on this line syslog is the source data type secondary is the workload set name the active flag is set and the number of qualifiers is 1 The second line of the workload set definition shows that System is the qualifier and it has one qualifier value F A blank line separates the list of workloads sets from the list of workloads Each workload consists of at least three lines with a blank line separating one workload from the next G The first line of each workload contains the source data type the workload set name the workload name and the workload sequence number syslog is the source data type primary is the workload set name sysjob is the workload and 001 indicates that this is the first workload in the primary workload set H The second line of the workload contains descriptive text describing the workload This text must be enclosed in single quotation marks For this example the text is not present I The third line of the workload begins the act
461. set parameter a TQ 01011 18 9 31 Customizing Data Collection Probes 9 9 4 9 32 Changing Default MAPPER Log File Names During the tqinit process at installation time you are asked to provide the file name for your MAPPER log files This information is stored in the performance database and used by tqmpr for selecting the MAPPER log files for processing If you specify a cycle number tqmpr selects all cycles of the MAPPER log file greater than the specified cycle If you do not specify a cycle number tqmpr selects all cycles found on the system The cycle number of the last file processed is then written to the database This allows tqmpr to process all current cycles of the MAPPER log file automatically To change the file name you must specify one additional parameter j on the tqgmpr processor call statement See 9 9 2 for the format of the processor call statement The tqmpr program stores the new file name and its cycle number in the performance database The tqgmpr program uses the new information for selecting files for processing The format of this additional parameter is as follows j acctng log n where j specifies the latest cycle of the MAPPER accounting log file name that has been processed acctng log is the name of the MAPPER accounting log file n is the file cycle number tqmpr will process all the cycles currently found on the system starting with cycle n 1 Note The j parameter and its parameter
462. shold checking is done at each sample the number of seconds specified is actually converted into the number of samples The tqzip program is used to define threshold information for a parameter When defining thresholds for a parameter you can define either two or three threshold zones If you have two threshold zones you specify only one threshold value If you have three threshold zones you specify two threshold values Each threshold value is the point that separates one threshold zone from another Alarm messages generated by TeamQuest Baseline include a severity indicator The following severities are used for the parameters with two threshold zones NORMAL The value for the parameter has returned to the normal zone CRITICAL The value for the parameter has crossed the threshold leaving the normal zone The following severities are used for parameters with three threshold zones NORMAL The value for the parameter has returned to the normal zone WARNING The normal zone is zone 1 or zone 3 and the value for the parameter has moved one zone away from the normal zone CRITICAL The value for the parameter has moved two zones away from the normal zone or The normal zone is zone 2 and the value for the parameter has moved one zone away from the normal zone 7 2 TQ 01011 18 Customizing Alarms and Thresholds Figure 7 1 illustrates the concept of threshold zones At each data sample the probe performs threshold checki
463. should appear on the second and subsequent lines of the processor call TQ 01011 18 Customizing Data Collection Probes The parameters specified on the processor call statement must be separated by commas Do not split single parameters across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keyword The tqodtp probe has a default collection period retention period and sample interval See Section 11 for more information on maintaining table data The defaults are Collection period The collection period indicates a time range that the tqodtp probe when running is to sample system activity By default tqodtp will continuously collect data Sample interval The sample interval in seconds indicates how often the tqodtp probe is to sample system activity within a collection period The default sample interval is 1 minute Retention period A retention period represents how long the data is maintained in the database A retention period called Primary exists for each table and has a default retention period of 24 hours 9 6 2 Terminating Data Collection To terminate the tqodtp run from a demand run that has full console mode capabilities or greater enter CONS TQODTP END To terminate the tqodtp run from the system console enter TQODTP END 9 6 3 Abnormal Termination of Data Collection If tqodtp terminates in error due to a maximum time
464. source Subcategory Statistic Name Reqs Sec KWords Req Hit Rate Transfer Time RET FLD Hardware Serv Time Speculation Success Current NSRI Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req Transfer Time RET FLD Hardware Serv Time of All Reqs Category Group Category Resource Subcategory Statistic Name Reqs Sec KWords Req TQ 01011 18 Logical channel hit rate Logical channel transfer time Logical channel service time Logical channel hardware service time Logical channel concurrency IO Activity Logical Channel LCO LC1 Cache Summary Logical channel request per second Logical channel K words per request Logical channel hit rate Logical channel transfer time Logical channel service time Logical channel hardware service time Logical channel speculation success Current value of number of segments rolled in NSRI IO Activity Logical Channel LCO LC1 Read Hit Logical channel read hit requests per second Logical channel read hit K words per request Logical channel read hit transfer time Logical channel read hit service time Logical channel read hit hardware service time Logical channel read hit percent of all requests IO Activity Logical Channel LCO LC1 Write Hit Logical channel write hit requests per second Logical channel write hit K words per request A 25 OS 2200 Parameters Transfer Time RET FLD Hardware Serv Time of Al
465. spaces the quotation marks are not needed TableTitle row col text A title that appears on the first page where data from this table is printed The table title does not appear on subsequent pages of the same table The text must be enclosed in double quotation marks The row col is optional and indicates the row and column on the page where the text string should start The parentheses must be included in the row col specification You can specify the row or column number in one of two ways To indicate the absolute row or column specify only a number for the row and column To indicate a relative position the number should be preceded by a plus sign The relative position is calculated from the position where the previous character was printed on the page Optionally you can indicate just a row specification by row In this case the text is centered on the page You can specify more than one TableTitle key TextFields row col text A text string containing informational text that is displayed at the beginning of the table section of the report The text must be enclosed in double quotation marks The row col is optional and indicates the row and column on the page where the text string should start The parentheses must be included in the row col specification You can specify the row or column number in one of two ways To indicate the absolute row or column specify only a number for the row and c
466. spond to the capital letters in the preceding example The letters are not part of the actual screen display A The first line indicates that the data collection is ON and that 2 data collection periods are defined The column headings for the data collection periods indicate the number of the collection the start and end times the days of the week that are excluded from the data collection and the interval at which the data is collected if appropriate for the table The list of data collection periods shows the currently defined collection periods as well as periods when no data collection has been defined During the first collection period from 8 00 to 11 00 data is collected at 1 minute intervals During the second collection period from 13 00 to 16 00 data is collected at 30 second intervals For both collections data is not collected on Saturday or Sunday During the other periods of the day no data collection takes place TQ 01011 18 Customizing Tables 11 5 2 Adding a Collection Period The tqzip program allows you to add new collection periods for a table The following rules apply when adding a collection period A collection period can start at any time of the day and end at any point within 24 hours of the collection start time No collection period can overlap another collection period Collection intervals must be such that there are an even number of collection intervals in a day and they must be an ev
467. ss efficient to process since it contains data other than COD MIPS data For selection of the audit trail number see the n option in the format description The tqmips processor call statement specifies data collection parameters including input parameter table name database name and aggregation set names Normally you use the default values for these parameters as provided in the tgmips runstream To modify any of the defaults you need to update the tqgmips runstream in SYS LIB RUN The tqmips processor call has the following format tqmips r i parameter table L d database name L a aggsetl l a aggset2 L n audit trail number END TQ 01011 18 9 9 Customizing Data Collection Probes where Note executes tqmips in real time mode The account for the tgmips run must have a maximum real time level greater than zero The r option is recommended so that the tqmips probe has the best opportunity to report data without missing a sample as the data is being written to the COD audit trail and system log The default is to not run in real time specifies the location of the input parameter table name to be used by the tqmips probe For the tqmips probe the parameter table is the element SYS LIB TQ BASELINE MIPSINPUTTBL This is a required parameter on the tqmips call line specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system
468. sseeeeeeeenees 9 9 2 tqmpr Processor Call Statement ceeeees 9 9 3 MAPPER Log File Input eee eeeeeeeeeeeeteeeeeeeees 9 9 4 Changing Default MAPPER Log File Names _ 9 9 5 Workload Characteristics for the MAPPER Log Probe isinara ieee 9 10 System Log Data Collection oo cecceceeccesteeeeeesnees 9 10 1 tqsys Processor Call Statement eeeeees 9 10 2 System Log File Input eeeeeeeeeeeeeeeeeeeeeeeeeeeees 9 10 3 Workload Characteristics for the System Log Probe 9 11 User Data Probes sunia aenean chavjaheeeeetaaeeveeh 9 11 1 Writing a User Probe for Performance Data _ 9 11 1 1 Data Storage gt tiziac atest ie a 9 11 1 2 Running tqusrprb in Continuous Mode _ 9 11 1 3 Installing Your User Probe ou eee 9 11 1 4 Restrictions oo eeeeeeeeeeeeeeeeeeeeceeeeeeeteeeeeeeeeees 9 11 2 Writing a User Probe for Table Data _ 9 11 2 1 Creating a Table oo eeeeeeeceseeeseseeeeeteeeeeeeees 9 11 2 2 Storing Table Data 9 11 2 3 Examples of Table Data Collection 9 11 2 4 Command Syntax eeceeeseeeeeteeeeeeeeeees 9 11 2 5 Keyword Definitions eseese ere LLLLLLLLLLLLLLL ONDAMDAAMDAADNAOAAKRANH OO 9 28 9 29 9 31 9 32 9 33 9 33 9 34 9 35 9 37 9 38 9 39 9 45 9 46 9 47 9 47 9 48 9 49 9 57 9 58 9 62 9 65 XV Contents Section 10 Customizing Performance Databases 10 1 10 2 10 3 10
469. st be supplied when defining a derived statistic If a key is not actually used for a statistic null is displayed in the Derived Statistics edit box Many of the dialog boxes in derived statistic maintenance contain a statistic hierarchy list When the list is initially displayed the hierarchy is expanded to the first level that contains more than one item The first line of a statistic hierarchy list contains the name of the database you are working on Indented under the database name are the first level keys the category groups Indented under the category groups are the second level keys category Subsequent indented levels include the subcategory keys and statistic keys Each key item is placed on its own line in the list box To perform many of the derived statistic maintenance tasks you must expand the derived statistic list to the full hierarchy which includes the category group category subcategory and statistic To define a statistic fully the statistic key must be selected To expand a key item to the next level of the hierarchy select the item in the Derived Statistics list box and click the Expand button You can also double click the item in this list box to display all of the items in the next level of the hierarchy TQ 01011 18 Customizing Derived Statistics The following dialog box shows a derived statistics list fully expanded to show all key levels 4 Derived Statistic Maintenance Session cl01 Database
470. subsequent workloads from each other Loading Workload Definitions You can reload the performance database with a new or updated set of workload definitions by using the l and w options on the tqzip processor call You should use the file as created by the u and w options see 5 4 1 as input for loading workload definitions If you update the workload definition file you should ensure that the format of the file remains consistent with what was output by the u and w options Note The tqzip program deletes any existing workload definitions from your database before loading the new definitions Use the e option on tqzip if you do not want existing definitions deleted To load workload definitions with tqzip use the following processor call format tqzip d databasename 1 w workload file e where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 l specifies that a load operation is to be performed by tqzip The type of load operation is workloads as indicated by the w option w loads the workload definitions into the database from the file specified by the workload file parameter The workload file is the name of a workload definition file created by the tqzip u and w options see 5 4 1 e loads the new workload definitions fr
471. sword See I 2 5 1 for information on using an encrypted userid and password runid xxxxx account xxxxx project xxxxx The run id account and project values are used on the RUN statement for the RSI run that is created to handle a socket connection Standard OS 2200 rules must be followed for each xxxxx value Account may also specify a user id for the RUN by entering account userid The account statement is required If running TQTCP in real time this account must allow real time mode Otherwise the RSI runs started to handle each connection will abort immediately Runid defaults to TQTCP and project defaults to TQDATA if they are not specified TQ 01011 18 I 5 2200 Socket Interface siteid xxxxx Optional list of site ids to use for RSI runs If not present site ids will automatically be generated by RSI Site ids are specified one per line Up to 50 concurrent RSI runs can be executing within TQTCP Each socket connection initiates an RSI run If using specific site ids TQTCP will use the first 50 site ids specified here If there are more the additional ones are ignored If there are fewer the number of site ids specified is used as the maximum number of concurrent RSI runs xqtfile qualifier filename This file is where the tqdata executable resides This file is used in the RSI run to execute tqdata If this statement is not supplied the default is SYS LIB TQ BASELINE The following is an example of a TQTCP c
472. t Online only report definitions in the SIP set of directories will have meaningful results SIP Reports The SIP reports described in this subsection provide detailed information about selected areas of system performance I O Reports The I O report definitions are found in the io subdirectory of the sip subdirectory File Name Description File Name Description File Name Description File Name Description File Name Description CU UNIT RPT Activity by Control Unit Gives detailed statistics for the selected control units CU TOPN RPT Most Active Control Units Gives detailed statistics for the top n control units CD DKPER RPT Cache Disk Percent Provides a summary of cache actions RLP RATE RPT RLP Rates Shows record lock processor RLP rates MH REQ RPT Multihost Reqs Sec Provides system wide requests per second information on Multihost usage Parameters presented are Lock Reqs Sec Unlock Reqs Sec Delayed Resp Msg Sec and Deadlocks Sec TQ 01011 18 Predefined Report Descriptions File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description TQ 01011 18 MHTIPREQ RPT Multihost TIP Reqs Sec Provides total TIP requests per second information on Multihost usage Parameters p
473. t View for data display processing TQ 01011 18 1 5 Introduction 1 5 1 1 5 2 Data Collection TeamQuest Baseline and TeamQuest Online use system monitors in the OS 2200 Executive Exec to gather data TeamQuest Baseline and TeamQuest Online probes collect data from the various system monitors and store that data in the performance database This data is then available for future retrieval and reporting The performance database holds all data used to generate reports and to monitor the system in real time Both TeamQuest Baseline and TeamQuest Online use the performance database for data storage Data Display TeamQuest View the tqrpt program and the tqrep program can be used to display the data collected and stored in the performance database TeamQuest View is a graphical user interface providing graphical representation of data The tqrpt program provides character base reporting for aggregation set data see Appendix D The tqrep program provides character base reporting for table data see Appendix E During the data display phase the TeamQuest View software retrieves data from the performance database You can then use this data to generate reports and display them in textual or graphic format See the TeamQuest View User Guide for more information Figure 1 2 shows the components of TeamQuest Baseline TeamQuest Online consists of all these components except the workload probes which receive data from HVTIP TPM IO
474. t to the contents of User Text 3 if that field did not contain this SNMPtTrap command Database This is set to the name of the database that contains the data used to detect the alarm condition This field can be used to open a report for a specified parameter using the TeamQuest View command line Parameter This is set to the parameter name of the entity that raised the alarm This field can be used to open a report for the specified parameter using the TeamQuest View command line These TeamQuest Baseline specific fields are part of the TeamQuest Enterprise MIB see J 6 This MIB is supplied with TeamQuest Baseline in text form so that it can be loaded into a management console if desired The MIB defines a group called tq2200AlarmTrap that contains the variable trap fields described above The field names in the MIB have tq2200Alarm as a prefix on the field names Management Console Configurations The TeamQuest Baseline comes with configuration information for the Hewlett Packard OpenView product and the Tivoli Enterprise Console TEC product The configuration information supplied for the HP OpenView product requires that at a minimum the System Identifier Severity Value and ThreshZone trap variables be sent with the TeamQuest Performance Software alarm trap For the TEC product the System Identifier and Value trap variables must be sent with the trap You can use the field list on the SNMPTrap command to specify which tr
475. table Once the table is created you are ready to build the runstream to perform the data collection and storage 9 56 TQ 01011 18 Customizing Data Collection Probes 9 11 2 2 Storing Table Data Data for a sample is contained on a series of input lines and is separated from other samples by a blank line Each sample of data contains two parts one containing the group key values and the other containing the records that describe the entities being measured in the sample interval The group keys are contained on the first line of the sample input set Each record is contained on a line and these lines follow the group key line The following shows an example 01 04 12 14 00 30 1800 CLO1 A 123456 7 20345 B acc001 117 129678 J5KH2 246 345123 SYSTEM 520 975123 lt blank line gt The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A This line specifies the Group key values for the sample Group key fields are the fields that describe a sample interval and do not change across all of the records in that sample This line consists of the Time Interval and System fields from the table specification In this example the user provided the date time 01 04 12 14 00 30 and the system CL01 on the input line The value of 1800 indicates a desired sample interval of 30 minutes represented in seconds 60 seconds per minute 30 minutes
476. tation of a report e Parameter Group these sections define the data series that are included in the report One parameter group section appears for each data series in the report C 2 TQ 01011 18 Report Definition File C 4 General Section The general section contains the specifications for the layout of the report Format The format for the general section is general Copyright Copyright c 2012 TeamQuest Corporation All Rights Reserved Product Name text Report Name text Host Type host type Date Format format Time Format format Aggregation Set aggregation set name Report Type Monitor History Date Multiplier positive integer Point Size positive integer Time start end Date CalendarType RelativeDate RelativeDate View Spread Sbar Pie Cluster Line Area Point Scatter Trend Format XAxis LegendEntries Page Trend Series Line Order Equation Scatter XSeries YSeries Line Order Equation XAxis Title text Value Title text Title Style Font Style Size Foreground Background Border Value Style Font Style Size Foreground Background Border Value Axis Enable Font Style Size Color Value Line Enable Thick MajorTick MinorTick MajorGrid MinorGrid Values Major Minor Minimum Maximum Metric XAxis Style Font Style Size Foreground Background Border XAxis Enable Font Style Size Color XAxis Line Enabl
477. tatistic The statistic name is limited to 51 characters in length and is case sensitive Statistic Hierarchy displays the hierarchy for the database you are working on The status box at the bottom of the list displays the type of key name that is selected in the list Use the Statistic Hierarchy box to specify where the newly derived statistic is placed in the database hierarchy Derived statistics can only be placed at the statistic level within the database hierarchy Therefore the item selected must be a subcategory or a statistic You must have entered at least one character in the Statistic edit box and have selected a subcategory or statistic item in the Statistics Hierarchy list box before the OK button is enabled These two items together determine where the newly derived statistic is placed in the statistic hierarchy When you click OK the new derived statistic is placed in the statistic hierarchy You must click OK in the Derived Statistics Maintenance window for the newly derived statistic definition to be sent to the database Expand button allows you to expand the hierarchy list of a selected key item to the next level View button allows you to see the key values that are truncated in the list box see 8 3 5 8 3 5 Viewing Long Key Names Many of the dialog boxes contain a View button which allows you to view complete key name fields that have may be truncated in the dialog box When you click the View button the Vie
478. tatistics This section contains advanced concepts for using derived statistics with TeamQuest Baseline It also tells you how to use the tqzip program to unload and load your derived statistic definitions to an output file Section 9 Customizing Data Collection Probes This section provides information you need to customize the data collection for each of the TeamQuest probes It also tells you how to write your own probes for collecting site specific data Section 10 Customizing Performance Databases This section tells you how create additional databases delete databases estimate disk space requirements for a database clear database locks and remove database parameters from a database It also provides backup and recovery information Section 11 Customizing Tables This section provides the information you need to maintain the table data collected by the TeamQuest Baseline probes It tells you how to use the tqzip program to define collection periods and retention specifications delete records from tables and register a table in the database It also tells you how and why to use the tqrorg program to reorganize your table files for improved efficiency TQ 01011 18 v About This Guide Appendix A OS 2200 Parameters This appendix lists the statistics that are collected by the TeamQuest Baseline and TeamQuest Online host software Appendix B Predefined Report Descriptions This appendix provides a description of each of th
479. te the derived statistic definition file you should ensure that the format of the file remains consistent with the derived statistic definition file created by using the u and s options see 8 4 1 Notes e Ifyou are using derived statistics for the workload performance reports you will need to load the derived statistics before data values are available See A 8 for information on workload performance derived statistics e When loading derived statistic definitions the tqzip program deletes any existing derived statistic definitions that are not contained in the file being loaded Use the e option on tqzip if you do not want existing definitions deleted To load derived statistic definitions with tqzip use the following processor call format tqzip d databasename 1 s derived file e where d specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 l specifies that a load operation is to be performed by tqzip The type of load operation is derived statistics as indicated by the s option S loads the derived statistics definitions into the database from the file specified by the derived file parameter The derived file is the name of a derived statistics definition file and must be in the same format as the file produced by the u and s o
480. te tqdata If this statement is not supplied the default is SYS LIB TQ BASELINE TQ 01011 18 7 2200 Socket Interface The following information should be entered for each user allowed to run TQ View or TQ Administration userid xxxxx The user id and password values are used to initiate the RSI run that is created to handle a socket connection These statements are required and standard OS 2200 rules must be followed for each xxxxx value The user id must have the security attribute specifying that the RUN image is always user entered or RUN image is user entered or system generated Also terminal timeout must be disabled for the user id If you need to use an encrypted user id and password use the euser and epswd statements euser xxxxx An encrypted user id euser that can be substituted for the user id See 1 2 5 1 for information on using an encrypted user id runid xxxxx account xxxxx project xxxxx The run id account and project values are used on the RUN statement for the RSI run that is created to handle a socket connection Standard OS 2200 rules must be followed for each xxxxx value Account may also specify a user id for the RUN by entering account user id The account statement is required If running TQTCP in real time this account must allow real time mode Otherwise the RSI runs started to handle each connection will abort immediately If not supplied the run id defaults to the first 6 characters of the
481. tenance tasks can be performed through this dialog box The session name and the database for which aggregation set maintenance is being performed are listed at the top of the dialog box The Aggregation Sets list box contains the list of aggregation sets defined for the database This dialog box is displayed with an aggregation set automatically selected All operations are performed relative to the selected aggregation set The Open button opens the selected aggregation set The New button inserts an aggregation set after the selected item The Delete button deletes the selected aggregation set Select an aggregation set from the list box before performing any of the operations with the Open New or Delete buttons When you click OK your changes are made to the aggregation set definitions including any changes made to aggregation set attributes accessed through the Open or New buttons see 6 3 2 and 6 3 3 When you click Cancel any changes you have made are ignored TQ 01011 18 6 9 Customizing Aggregation Sets 6 3 2 Creating an Aggregation Set To create an aggregation set definition click the New button in the Aggregation Set Maintenance dialog box see Performing Aggregation Set Administration Tasks6 3 1 The Aggregation Set Attributes dialog box is displayed You define a new aggregation using this dialog box Aggregation Set Attributes Session cl0l Database PRODUCTION OK ok Cancel e Help Aga Set Na
482. ter name If the parameter list is longer than will fit on one screen enter a blank line to view more of the list After you select a parameter name the tqzip program displays the threshold definition for that parameter If you enter y at the final prompt the tqzip program prompts you for another parameter name to view If you enter n at the prompt the tqzip program returns to the Alarms and Thresholds Menu see 7 3 Creating Derived Statistic Threshold Definitions To create a threshold definition for a derived statistic select Define Thresholds on the Alarms and Threshold Menu see 7 3 by entering 1 at the prompt The tqzip program displays the following prompts to define thresholds for a derived statistic gt Enter your selection gt 1 gt gt Create threshold Update threshold or Exit c u lt e gt A gt C gt gt System Name B gt 1 DERIVED gt 2 venus gt gt Enter number name lt blank gt for more or E to exit gt gt gt Statistic Name C gt 1 Derived Statistic gt 2 RespTimeGoal gt 3 ResponseTime gt 4 UserCpuutil aoe number name lt blank gt for more or E to exit 3 gt Select the resource s to use for this derived statistic D AR Name precede choice with to deselect TQ 01011 18 Customizing Alarms and Thresholds gt 1 IP000 gt 2 IP001 gt 3 IP002 gt 4 IP003 gt gt Enter number name lt blank gt for more or E to e
483. ters collected by the TeamQuest Baseline MAPPER log probe This probe collects data based on the defined workload sets Note These parameters are not collected by TeamQuest Online The following statistics are collected by workload set workload category group category subcategory and statistic Parameter Hierarchy Category Group Category Subcategory Statistic Name I Os Processed Lines Processed Breakpoint Run Functions Display Functions Connect Time Active Time ms Avg Sec Response Functions Logged Non Run Funcs Logged Functions Started Funcs Under Run Logged Funcs Under Run Started Run Lines Processed Run I Os Run Read I Os Interactions Response Sec Int Run Interactions TQ 01011 18 Mapper Log Workload Total Number of input and output operations processed Number of lines read and written Number of breakpoints that occurred Number of run functions that occurred Number of display functions that occurred Total time in seconds that stations were connected Total function active time in milliseconds Number of seconds per logged function Number of logged functions that occurred Number of logged functions that occurred outside of runs Number of MAPPER functions that were processed Number of logged functions that occurred under runs Number of started functions that occurred under runs Number of lines read and written from functions under runs Number of input and output operations from fun
484. th TeamQuest Baseline 7 32 SIP data collection description 9 2 file input totqbsp 9 4 starting 9 2 terminating 9 3 SIP parameters category groups A 2 dedicated Java processor statistics A 14 T O activity A 24 memory statistics A 18 paging statistics A 22 performance monitor statistics A 9 processor statistics A 7 program load statistics A 18 queue statistics A 39 system statistics A 5 XPC statistics A 36 SIP reports T O reports B 2 memory reports B 7 processor reports B 6 B 8 summary reports B 11 system reports B 10 SNMP traps alarm administration J 2 CMS 1100 configuration J 2 Computer Associates configuration J 9 content J 4 HP OpenView configuration J 7 management console configurations J 6 mappings J 4 Index 5 Index overview J 1 sending J 1 special considerations J 12 syntax J 2 TeamQuest MIB definition J 9 Tivoli configuration J 7 variable fields J 5 socket connection defining 4 5 sockets connections I 11 defining connections 4 1 interface overview I 1 special considerations for SNMP traps J 12 specifying workload definitions 5 1 SSL connection defining 4 6 storing table data 9 57 summary reports B 11 syntax of SNMP traps J 2 system reports B 10 statistics A 5 System Activity Monitor Runs table information A 58 Transactions table information A 59 System Activity Monitor data collection description 9 5 parameters A 58 starting 9 5 terminating 9 6 system log data
485. th a value of 0 Team Quest Baseline will report zeros in the display of this data Program Load Statistics The following statistics are available for Exec level SB4R3 and higher Parameter Hierarchy Category Group Memory Category Program Load Subcategory Total Statistic Name Init Program Size Average initial program size in K words Init Program Loads Total initial program loads per second Sec Init Bank Size Average initial bank size in K words Init Bank Loads Sec Total initial bank loads per second Dyn Bank Size Average dynamic bank size in K words Dyn Bank Loads Sec Total dynamic bank loads per second Category Group Memory Category Program Load Subcategory Demand Abs Statistic Name Init Program Size Average initial program size in K words Init Program Loads Total initial program loads per second Sec Init Bank Size Average initial bank size in K words TQ 01011 18 OS 2200 Parameters Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init Bank Size Init Bank Loads Sec Dyn Bank Size Dyn Bank Loads Sec Category Group Category Subcategory Statistic Name Init Program Size Init Program Loads Sec Init B
486. the data obtained at each sample interval to determine if the data values exceed user specified threshold limits If a data value for a parameter is found to exceed a threshold the probe can generate an alarm to alert you that a threshold condition has been met To determine when a threshold condition has been met the TeamQuest Baseline probes use the concept of threshold zones You define threshold information for a parameter including the number of zones the normal zone and the number of seconds before the alarm is triggered The probe generates an alarm whenever the value for a given parameter is found to be out of the current zone for the specified number of seconds TQ 01011 18 7 1 Customizing Alarms and Thresholds The following terms are important to the understanding of threshold zones e The number of zones defines the number of threshold zones in use for a given parameter It is possible to have either two or three threshold zones for a parameter e The normal zone is the zone that the parameter value normally falls into e The current zone is the zone that contained the last value for the parameter When an alarm is generated the current zone is changed to be the zone that contains the current value for the parameter The current zone is initially set to the normal zone e The seconds before alarm specifies the number of seconds that the parameter value must be out of the current zone before an alarm is raised Since thre
487. the aggregation set definition and all data contained in the aggregation set are deleted When aggregation sets are loaded only the aggregation set definition is created in the database No data is contained in the loaded aggregation sets until either the TeamQuest Baseline probes or the tgagg program store data into the aggregation set To load aggregation set definitions with tqasa use the following processor call format tqasa d databasename 1 aggset file r ALLSETS where d specifies the database for which you want to load aggregation set definitions The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 l specifies that an aggregation set load operation is to be performed by tqasa All aggregation set definitions in the file specified by the aggset file parameter are imported into the database r ALLSETS deletes all existing aggregation sets before loading the new ones If you do not specify this option no existing aggregation sets are deleted 6 24 TQ 01011 18 Customizing Aggregation Sets Example The following shows an example of loading aggregation set definitions contained in the file aggsetfile Because the r ALLSETS option is specified all existing aggregation sets will be deleted before the new ones are loaded gt tqasa 1 aggsetfile r ALLSETS gt TQASA 7R4G 0
488. the performance database When an alarm condition is encountered the probe generates an alarm Alarms are handled by the tqalm program that is provided with TeamQuest Baseline The tqalm program can perform the following tasks for each alarm e Send an alarm message to the system console e Calla user provided routine for site specific alarm handling e Optionally send an SNMP trap to a network management console see Appendix J In addition to alarm handling the tqalm program performs threshold checking for all derived statistics that have a threshold definition in the performance database Every n seconds where n is the granularity of the aggregation set used by tqalm tqalm retrieves the data needed to calculate derived statistic values performs threshold checking on the values and generates alarms as needed To allow tqalm to do threshold checking using the most current data available it is important that it uses the same database and finest granularity aggregation set that is being used by the data collection probes running continuously on your system that is tqbsp Use the d and a options the tqalm processor call to specify the database and aggregation set The tqalm program must be running on your system in order for any alarm handling to occur To start tqalm use the tqalm runstream in SYS LIB RUN The tqalm runstream is started automatically when the TeamQuest Baseline control runstream tqctl is started See 9 2 for more info
489. tion are displayed in the status bar at the bottom of the main window Each time you define a new session the host name or IP address is added to the list of currently defined hosts 4 3 TeamQuest Performance Software Administration New Session Dialog Box The New Session dialog box contains the following fields Connection These option buttons allow you to select the type of connection to be used to access the performance database The fields on the right side of the dialog box will change depending on the type of connection you select in this field Socket Settings These fields are used to specify the settings for socket connections Host Name or IP Address This field contains the name of the host or the IP address of the system where the performance database resides Port This field contains the port identification for the socket to connect to Socket name list This list box contains a list of the socket names that have already been used with a session Each time you define a new session the host name or IP address is added to this list For more information see 4 4 1 SSL Settings These fields are used to specify the settings for SSL connections Host Name or IP Address This field contains the name of the host or the IP address of the system where the performance database resides Port This field contains the port identification for the socket to connect to Socket name list This list box contains a list of t
490. tion is in effect The default values for Point Consolidation are defined in the Miscellaneous settings of the User Preferences See the TeamQuest View User Guide for information on User Preferences Legend Enable Location top left bottom right TQ 01011 18 The specifications determining where a legend appears on the report Enable Indicates whether the legend appears on the report The acceptable values are ON or OFF If Enable is set to ON the legend appears on the report If Enable is set to OFF the legend does not appear on the report Location Specifies the location of the legend on the report Right Places the legend to the right of the chart Left Places the legend to the left of the chart Top Places the legend above the chart Bottom Places the legend below the chart Float Places the legend over the top of the chart Report Definition File If Location is float the fields top right left and bottom are required to place the legend properly If the location is not float the fields top right left and bottom are ignored top Specifies the relative location for the upper coordinate of the top left hand corner of the window containing the legend left Specifies the relative location for the left coordinate of the top left hand corner of the window containing the legend bottom Specifies the relative location for the lower coordinate of the bottom right hand corner of the window containing the le
491. tion set to receive data from another aggregation set in the database If you do not specify a value the aggregation set receives data only from TeamQuest Baseline probes Note that TeamQuest Baseline probes only support Group_Types of DAY G Comment lines begin with a character D This line begins the second aggregation set definition E The last line of the file must be the END AGGSETS line TQ 01011 18 6 23 Customizing Aggregation Sets 6 4 5 Loading Aggregation Set Definitions You can reload the performance database with a new set of aggregation set definitions by using the l option on the tqasa processor call You should use the file as created by the tqasa u option as input for loading aggregation set definitions see 6 4 4 If you update the aggregation set definition file you should ensure that the format of the file remains consistent with what was output by the tqasa u option Existing aggregation set definitions cannot be updated by loading a new set of definitions If the load process encounters an existing aggregation set a warning message is printed and that definition is ignored If you want to replace your existing aggregation sets with the ones in the load file use the r ALLSETS option with the l1 option This will cause all existing aggregation sets to be deleted before the new ones are loaded Note Use of the r ALLSETS option will cause all existing aggregation sets to be deleted from your database Both
492. tions Application Program Interface error status and aux is the Communications Application Program Interface auxiliary status See the ClearPath Enterprise Servers Communications Application Program Interface COMAPD User s Guide for a description of these error statuses TCP_xxxxx is the name of the Communications Application Program Interface function that was called when the error occurred for example TCP_LISTEN TCP_SEND or TCP_RECEIVE connection_id is the session identifier obtained from the Communications Application Program Interface for the TCP session that encountered the error TQ 01011 18 I 9 2200 Socket Interface 1 2 5 Security Considerations 1 2 5 1 Since the TQTCP configuration element contains a valid user id password and account you may want to limit access to this information You can use OS 2200 file security to limit access to the configuration element To do this 1 Catalog a new file and copy SYS LIB TQ BASELINE TQTCP CONFIG to it 2 Attach the appropriate security to this new file using SIMAN 3 Update the new configuration element to include at a minimum the user id password and account configuration statements 4 Update the TQTCP processor call in the TQTCP runstream to include the c option to specify the location of the secured configuration element Remember that at a minimum the TQTCP run must be able to read the configuration element User id and Password Encryption A
493. to prompt you to modify Bottom N settings LastPoint If set to On only the last point is used to determine the Bottom N order If set to Off or not set the entire report calendar range is used to determine the Bottom N order Include the reserved keyword to cause the interface to prompt you to modify Bottom N settings C 6 Individual Parameter Setting Sections The following subsection defines threshold and multiplier settings on specific parameters or general settings that are applied to parameters with common key values These settings are applied to the parameters which result from the parameter groups defined in the report definition files Note These settings are overwritten when the report is saved The format for an individual parameter setting section is Parm id Keyname keyvalue Value Types Average StdDev Threshold Enablelow Lower EnableUp Upper Parameter On Off Display Multiplier Lower Threshold Process executable name Upper Threshold Process executable name C 32 TQ 01011 18 Report Definition File The following descriptions explain the syntax of the individual parameter setting specifications of the report definition Parm id This is a unique name for an individual parameter setting The individual parameter setting can be a string of up to 24 characters The name must begin with Parm and id is a numerical value which makes it unique from other individual para
494. to the Extended Processing Architecture of OS 2200 computer systems This includes the 2200 500 and the 2200 900 systems XPC Abbreviation for the Extended Processing Complex This processing complex can be used to accelerate file accesses Glossary 6 TQ 01011 18 Bibliography Business Information Server for ClearPath OS 2200 Administration Guide 7831 9415 Unisys Corporation ClearPath Enterprise Servers Communications Application Program Interface COMAPD User s Guide 7851 5749 Unisys Corporation ClearPath OS 2200 Cipher Application Programming Interface API Programming Reference Manual 3826 6110 Unisys Corporation ClearPath OS 2200 Exec System Software Administration Reference Manual 7831 0323 Unisys Corporation ClearPath OS 2200 Software Products Installation Guide 7831 0612 Unisys Corporation ClearPath OS 2200 TeamQuest PMLog End Use Reference Manual TQ 01351 TeamQuest Corporation ClearPath OS 2200 TeamQuest Probes End Use Reference Manual TQ 01361 TeamQuest Corporation ClearPath OS 2200 TeamQuest Products Installation Guide TQ 01013 TeamQuest Corporation ClearPath OS 2200 Transaction Processing Administration and Operations Reference Manual 7830 7881 Unisys Corporation Distributed Processing Middleware Open Distributed Transaction Processing Administration Guide Volume 1 Installation Configuration and Ongoing Administration 7833 5072 Unisys Corporation OS 2200 Co
495. tor option UTIL SUM RPT Processor Utilization Provides processor utilization statistics for user categories SW UNIT RPT Proc Util by Unit Shows detailed processor utilization by switching level information for the selected processors The statistics displayed on this report are created when SIP level 4 or higher is configured SW SUM RPT Processor Utilization Shows summary processor utilization by switching level information for the system The statistics displayed on this report are created when SIP level 4 or higher is configured IDLE RPT Processor Utilization Idle Provides processor utilization statistics for idle as well as the minimum and maximum idle for the time range UTIL TIP RPT Processor Utilization TIP Provides TIP processor utilization as well as the minimum and maximum values for the time range Predefined Report Descriptions B 3 5 System Reports B 10 The system report definitions are found in the system subdirectory of the sip subdirectory File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description File Name Description BAT RUNS RPT Batch Run Summary Summarizes batch population and response information CAL TOPN RPT Most active CALLs Used Shows the n most act
496. trail number END where u processes the user area of TPM termination records using the LINC format described in 9 7 6 Z is used to register the tqtpm probe in the database i specifies the location of the input parameter table name to be used by the tqtpm probe For the tqtpm probe the parameter table is the element SYS LIB TQ BASELINE TPMINPUTTBL This is a required parameter on the tqtpm call line d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini element is used Make sure that all programs called in a runstream use the same database name a specifies the aggregation set to be used for data collection The aggset1 parameter is the name of an aggregation set defined in the database name The aggset1 parameter is case sensitive The aggregation set is optional on the call line If no aggregation set is specified the Aggset1 setting of the Aggregation Sets section in the tq ini element is used See 2 2 1 for a description of aggregation sets n specifies the TPM audit trail number By default this value is 17 corresponding to SYS ACI 17 and the SYS F017L1 audit trail file names used in file selection Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These
497. tries define the Unisys OS2200 TeamQuest enterprise number to trapd and format a message for the event browser xnmevents for each trap type Additionally a pop up window is displayed for critical alarm traps Two symbolic elements have been supplied to assist you in updating the HP OpenView trapd conf file These elements are located in the file SYS LIB TQ BASELINE The element names are OV3TRAPD and OV4TRAPD You should use the element that corresponds to your level of HP OpenView for example HP OpenView level 3 users would use OV3TRAPD and HP OpenView level 4 or 5 users would use OV4TRAPD 3 When you have updated the trapd conf file you must follow the HP OpenView procedures for restarting the HP OpenView trap daemon J 5 2 Tivoli Management Environment Configuration To make your TeamQuest Performance Software alarm display a message on a TEC Event window you must first define a threshold using the tqzip program configure the TEC SNMP Event Adapter to accept and process the alarm trap and define the TeamQuest event classes to the TEC Event Server Follow these steps 1 Define a threshold using the tqzip program and add the SNMPTrap command to one of the User Text fields in the threshold definition Since you can select fields in the TEC SNMP Event Adapter configuration you do not need to specify which fields to send on the SNMPTrap command On the SNMP MGMT statement in the CMS 1100 configuration you must specify the h
498. ts system server programs and service routines are collected The tqodtp program uses a monitoring interface supplied with the Unisys OLTP TM2200 product ODTP data is monitored by the tqodtp program at every sample interval and data is stored in several tables See Sections 2 and 11 for more information on the collection of table data tqodtp Processor Call Statement The tqodtp runstream contains the tqodtp processor call statement that controls the collection of information on active client programs server programs and service routines in an ODTP environment The tqodtp processor call has the following format tqodtp r L d database name END where r executes tqodtp in real time mode The account for the tqodtp run must have a maximum real time level greater than zero By running in real time mode you can prevent the probe from missing data as a result of being swapped out during scheduled sample periods The default value is to not run in real time mode d specifies the name of the database to be used for data collection The database name parameter is the name of an existing database on your system If this parameter is not specified the Database setting in the tq ini file is used Be sure that all programs called in a runstream use the same database name Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters
499. ts Sunday the second represents Monday and so on through the remaining days of the week The last digit represents Saturday For example if you want your aggregation set to exclude Sunday and Saturday and to include all other days the exclude parameter is specified as 1000001 If you want the aggregation set to include only Sunday and Saturday the exclude parameter is specified as 0111110 If you do not specify the exclude parameter 0000000 is assumed that is no days are excluded from the aggregation set Customizing Aggregation Sets w specifies the day that is the first day of the week for an aggregation set with a group type of week The weekstart parameter can be a digit from 0 to 6 The digits represent the days of the week as follows 0 Sunday 1 Monday 2 Tuesday 3 Wednesday 4 Thursday 5 Friday 6 Saturday If you do not specify the weekstart parameter 0 is assumed p specifies that the aggregation set to be created is in expanded format See 6 2 5 for a description of compact and expanded aggregation sets If the p option is not used the aggregation set is created in compact format Example The following example creates a compact aggregation set called 30 minute with a granularity of 30 minutes and a retention period of 35 days gt tqasa i 30 minute g 30min k 35 gt TQASA 7R4G 011312 11 04 45 Fri Jan 13 14 49 08 2012 gt Copyright c 1992 2012 TeamQuest Corporation All Rights Reserve
500. tton to remove the workload definitions New workload sets and workload definitions are inserted into the Workload Maintenance dialog box If a syntax error occurs the workload set and workload definitions are not inserted in the Workload Maintenance dialog box When importing new workloads with existing workload sets it is important to check the order of the workloads within the existing workload sets After an import function your updates can be deleted by clicking the Cancel button on the Workload Maintenance dialog box To accept the updates click OK and all updates are written to the session database See 5 4 1 for an example of a workload definition file Exporting Workload Sets and Workload Definitions To export all workload sets and workload definitions click the Export button on the Workload Maintenance dialog box see 5 3 1 When the Save File dialog box is displayed select or enter a file that will contain a workload definition for all workload sets and all workloads listed in the Workload Maintenance dialog box See 5 4 1 for an example of a workload definition file TQ 01011 18 5 17 Customizing Workloads 5 3 9 Deleting a Workload 5 4 5 18 To delete a workload from a database select the workload name from the Workload Info list box in the Workload Maintenance dialog box see 5 3 1 When you click the Delete button an alert box verifies your request to delete the workload When you click Yes in this
501. u can use a text editor to update the thresholds element before loading it into the database See 7 3 for more information on the threshold definition file format This section contains information you need to consider when updating the threshold element Threshold definitions can be updated and reloaded as many times as necessary Each time you load thresholds with tqzip the probes automatically begin using your latest threshold changes You can also use the tqzip interactive interface to maintain your threshold definitions See Section 7 for additional information on how to set up alarms and thresholds for your system 3 4 What s Next The following is a brief overview of the recommended steps that you can use to verify that the correct data is being collected after completing the basic installation of TeamQuest Baseline 1 Analyze the aggregation sets to ensure data is being collected in meaningful intervals By default TeamQuest Baseline collects data at 1 minute and 10 minute intervals These two aggregation sets feed the 1 hour and 8 hour aggregation sets for capacity planning and trend analysis Systems that are being monitored for performance purposes may require a longer data retention period for the 10 minute aggregation set while systems being monitored for capacity planning may require data retention longer than the 13 month default for the 8 hour aggregation set You can configure your aggregation sets for your site specific requ
502. ual workload definition specified in the workload definition language see 5 2 2 The workload definition can continue on subsequent lines and is terminated by a blank line Each line of the workload definition language must be enclosed in single quotation marks Notice that single quotation marks are also used within the braces to enclose text strings J This line begins the cmsmcb workload which is the second workload of the primary workload set Following the cmsmcb workload are the rest of the workloads in the primary workload set Then the workloads in the secondary workload set are listed TQ 01011 18 5 21 Customizing Workloads 5 4 2 5 22 The following general rules apply for the workload definition file e Each text field of the workload set and workload definitions must be enclosed in single quotation marks However text strings specified within the workload definition expression itself can be enclosed in single quotation marks or double quotation marks e Do not use quotation marks around the Active Inactive flag line B workload set qualifier count line B or workload sequence number line G e Workload sequence numbers must be sequential within a workload set and do not require leading zeros as seen in the preceding example see line G e Blank lines are significant and are used to separate the header from the first workload set definition the workload sets from the first workload and
503. uble Double Double String Double Double Double Double Integer Total SUPs accumulated for this transaction program since inception The current size in K words used by the transaction The type of transaction program being executed It can be one of the following self d self destructive self i self initializing reent reentrant onl bat online batch hv tip high volume TIP The time when the transaction program started The time this information was recorded in the database for this sample The actual time that this sample was taken for the transaction The desired sample interval in this collection period The system from which this sample was taken Total IP SUPs accumulated for this transaction program since inception Total I O SUPs accumulated for this transaction program since inception Total ER CC SUPs accumulated for this transaction program since inception Total I O requests accumulated for this transaction program since inception The number of combined occurrences TQ 01011 18 OS 2200 Parameters A 10 Open Distributed Transaction Processing Statistics This subsection contains a list of the data items collected by the TeamQuest Baseline Open Distributed Transaction Processing probe tqodtp The information collected by this probe is stored in the following tables System Table Information Server Table Information Client Table Information Service Table Information Note These data
504. udit control processing Symbionts Scaled percentage of Exec symbionts processing TQ 01011 18 A 51 OS 2200 Parameters Misc Switchable Scaled percentage of miscellaneous switchable Exec processing Real Time Scaled percentage of user real time processing ILES Scaled percentage of independently linked Exec subsystems CMS 1100 Scaled percentage of CMS 1100 processing UDS Scaled percentage of UDS processing TIP Scaled percentage of TIP processing Critical Deadline Scaled percentage of critical deadline processing Demand Scaled percentage of demand processing Batch Scaled percentage of batch processing Category Group Memory Category Paging Activity Subcategory Real Space Usage Statistic Name Memory Utilization Memory utilization for XPA systems A 52 TQ 01011 18 OS 2200 Parameters A 8 2 MAPPER Log Probe Derived Statistics The following statistics are collected by the MAPPER log tqmpr probe Parameter Hierarchy Category Group Category Subcategory Workload Set Workload Statistic Name IOs processed sec Lines processed sec Response sec display Throughput displays sec Population signed on Category Group Category Subcategory Workload Set Workload Statistic Name IOs processed sec Lines processed sec Response sec display Throughput displays sec Population signed on TQ 01011 18 Mapper Log Overall WLS1 WLS2 I Os processed per second Logic and data line
505. ue The field must be the name of a field that exists in the table The value can be either ON or OFF ON indicates that the TeamQuest Baseline probes will collect data for this field OFF indicates that no data collection will be done for this field If no value is specified for a field the FieldSpec line is ignored Fields that are part of the group key defined for the table cannot be unloaded or loaded If no FieldSpec line exists in the load file for a given table field the status of the data collection flag for that field remains unchanged in the table C Comment lines begin with a character D The last line of the file must be the END TABLESPECS line 11 11 Loading Table Specifications You can reload the performance database with a new set of table data collection and retention specifications by using the l and c options on the tqzip command line You should use the file as created by the tqzip u and c options as input for loading table specifications see 11 10 If you update the table specification file you should ensure that the format of the file remains consistent with what was output by the tqzip u and c options The tables named in the file being loaded must already exist in the database If a table does not exist a warning error is printed and the specifications for that table are ignored Existing table specifications for the tables being loaded are replaced with the data collection and retention informati
506. ue of the corresponding field in the data sample line see 9 11 1 1 is used to fill the resource component of the parameter hierarchy If no field is designated as a RESOURCE field the resource component of the parameter hierarchy is NULL NONE When a field is designated as NONE the value for the corresponding field is ignored The use of NONE is helpful when you want to skip an unwanted field produced by some commands The second line of the input specifies the weights for the values to be stored These weights tell the aggregator utility how to consolidate a set of data points from a finer granularity aggregation set into a single point in a coarser granularity aggregation set When you specify a weight you need to think about how you want to derive a single value from multiple samples The fields must be space or tab delimited You must specify a weight for each field in the first line of the input see A The aggregation weights must be one of the following SUM The coarser point is the sum of samples from the finer aggregation set AVG The coarser point is the average of samples from the finer aggregation set MAX The coarser point is the maximum of the finer values MIN The coarser point is the minimum of the finer values LAST The coarser point is the last value chronologically from the sequence of the finer values FIRST The coarser point is the first value chronologically from the sequence of the finer values statisti
507. up names If the derived statistic you chose in C requires qualifiers you are prompted for them now This example shows a derived statistic that requires system and resource name qualifiers A prompt did not appear for the system because there is only one system name other than DERIVED If only one qualifier exists in the prompt list it is automatically selected for you To add a selection to the qualifier list you must enter the number or name of the qualifier value To remove a selection from the list you enter the number of the qualifier value preceded by a hyphen An asterisk is used to indicate which qualifier values in the list have been selected You continue adding or removing selections in the qualifier list until entering an E to exit The tqalm program uses these qualifier values to determine how threshold checking is to be done for a derived statistic In this example the value of UserCpuUtil for IP000 and IP001 will each be calculated and separately checked against the UserCpuUtil threshold definition for possible threshold violations Threshold checking and alarm generation for IP000 and IP001 are independent of each other However any changes to the threshold definition for UserCpuUtil will affect both IP000 and IP001 After you have finished selecting the qualifier values for this derived statistic threshold the tqzip program displays the series of threshold definition prompts see G through P in 7 4 You can modify any
508. ure that all programs called in a runstream use the same database name TQ 01011 18 9 3 2 9 3 3 Customizing Data Collection Probes a specifies the aggregation set to be used for data collection The aggset1 and aggset2 parameters are the names of aggregation sets defined in the database name The aggset1 and aggset2 parameters are case sensitive All aggregation sets are optional on the call line If no aggregation sets are specified the Aggset1 and Aggset2 settings of the Aggregation Sets section in the tq ini element are used See 2 2 1 for a description of aggregation sets Note Parameters passed to a program from the Exec call line are in lowercase Some parameters on the program call statement are case sensitive These parameters should appear on the second and subsequent lines of the processor call The parameters specified on the processor call statement must be separated by commas Do not split a single parameter across two lines No continuation character is allowed when using multiple lines The processor call statement must be terminated with the END keyword Terminating Data Collection To terminate the tqbsp run from a demand run that has full console mode capabilities or greater enter CONS TQBSP END To terminate the tqbsp run from the system console enter TQBSP END Abnormal Termination of Data Collection If tqbsp terminates in error due to a maximum time contingency it restarts automatically If
509. us registration 11 19 Customizing Tables 11 9 Deregistering a Table 11 20 When a registered table is no longer needed it can be deregistered from TeamQuest Baseline When you deregister a table the table name is removed from the list of tables for the database Note Removing the table does not cause the table file to be removed but only the removes the table name from the table list You must remove the table file if you no longer need it To deregister a table select Table Registration from the Tables Menu see 11 3 by entering 5 at the prompt The tqzip program displays the following Table Registration Menu gt Enter your selection lt 1 gt gt 5 gt Table Registration Menu gt gt gt 1 Register a Table gt 2 Deregister a Table gt gt 99 Exit Table Registration Menu Enter 2 at the prompt to deregister a table The tqzip program displays the following prompts gt Enter your selection lt 1 gt gt 2 gt gt Registered tables A gt gt 1 2200 Archive Runs gt gt Enter Table Number B gt 1 gt gt Table 2200 Archive Runs deregistered The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A The tqzip program displays a list of the tables registered in the database This example shows one registered table 2200 Archive Runs B Enter the number of the table you w
510. user id and the project defaults to TQDATA The following is an example of a TQTCP configuration element using unique user ids interface CPCOMM port 2700 iF comment line include information for each user userid john runid john account 11000 project tqproj d comment line an encrypted user id for Sally euser Ofa3ca0a5 fb64 78f96e815528a4fb340614 account 11000 project tqproj Note Remember that the only required statements are userid and account If the defaults for the other configuration statements are acceptable you do not need to include any of the optional configuration statements in your TQTCP configuration element 8 TQ 01011 18 2200 Socket Interface In this example only userid john and sally are allowed to run TQ View or TQ Adminstration When a connection is made an RSI run would be started using the associated information above and the password solicited from either TQ View or TQ Administration The RSI runs for john and sally would look like the following john password run john 11000 tqproj sally password run sally 1100 tqproj 1 2 4 Error Messages If the TQTCP background run terminates prematurely error information will be available in the TQDATA TQTCP PRT file Communications Application Program Interface errors are output by TQTCP as follows Error COMAPI returned status aux on TCP_xxxxx for connection lt connection id gt where status is the Communica
511. usrprb program will store Each sample of data consists of one or more lines According to the command line options and the input the tqusrprb program creates parameters and stores values corresponding to those parameters in the performance database The fields must be space delimited Any warnings and errors are written to the run s print file Example 1 The following example shows how to store DMS AREA overflow record statistics in the performance database None of the TeamQuest Baseline probes collect this information and you may find it useful to have that information in your database The DMS DMU verify command provides this information To get this information into the performance database you can write a runstream and execute it to provide input for the tqusrprb program The DMS DMU verify command outputs information on the areas of a DMS database Database administrators are concerned as to when the overflow records are being filled This can signal the time for a database area expansion Statistics for the overflow records is part of the output from the DMU verify command This output can be captured into a file manipulated by an SSG skeleton and input into the performance database using tqusrprb Part of the output from the DMU verify command is as follows STATISTICS FOR AREA PRIMUS D01 CODE 301 TOTAL STATISTICS FOR AREA if PAGES TOTAL USED UNUSED STD UNUSED TOTAL WORDS PER PAGE PER PAGE DEV RECORDS PRIME 1000 17
512. ust specify a valid derived statistic name The descriptive lines must be in order from 1 through 8 All text in these lines must remain intact You can update the field values that is the values following the colons as long as you keep the fields in their same relative position on each line For example if you are modifying the descriptive information of line 3 do not alter the text Number of zones or Normal zone You can however change the field value for the number of zones from 3 to 2 The spaces surrounding field values are insignificant Lines 9 through 12 are for the qualifier values selected for this threshold definition The numbered line contains the qualifier name and number of values that follow The values are each on a separate line You can add additional threshold definitions by duplicating all the lines of a threshold definition and modifying the derived statistic name line and the descriptive information lines for the new threshold Remember to include the qualifiers needed to calculate the derived statistic values for threshold checking This is the first line of a threshold definition for a parameter Each parameter threshold definition consists of a parameter name line followed by eight lines of descriptive information The parameter name line must specify a valid parameter name The descriptive lines must be in order from 1 through 8 All text in these lines must remain intact You can update the f
513. ution is to take place Individual Substitution Group Substitution SYSTEM ALL_SYSTEM WLS ALL_WLS WL ALL_WL RES ALL_RES AG ALL_AG When using individual substitution the parameter reference is expanded into a set of individual parameter references using each of the values provided for the specified key The values used for substitution are obtained from either the report definition file or from prompting that is done when the report is opened The solving of the single derived statistic formula will actually report the values for the set of individual parameters For example if RES is used in a parameter reference for a disk statistic the formula is solved once for each disk specified for the report If four disks were selected when the report was opened the values for each of these four disks would be reported as a result of using this single derived statistic TQ 01011 18 8 2 3 Customizing Derived Statistics When using group substitution all the values for the specified key are used to expand out the list of arguments to a mathematical function Group substitution is only allowed in the argument list of a mathematical function The argument list is expanded as though a separate parameter reference was used specifying each key explicitly For example to find the average busy of all your disks you would use the AVG function and include a single parameter reference in its argument list The single parameter refe
514. val specified for the time range Drag the scroll box or use the arrows on the scroll bar to increase or decrease the size of the time interval The interval adjusts at the rate of the point size shown in the Granularity drop down list box If the granularity is less than 5 minutes 5 minutes is used When your definition is complete click OK The new aggregation set name is inserted into the Aggregation Sets list box in the Aggregation Set Maintenance dialog box see 6 3 1 To define the aggregation set to the host software you must click OK in the Aggregation Set Maintenance dialog box Note TQ 01011 18 When you create an aggregation set a new database file is created Customizing Aggregation Sets 6 3 3 Displaying or Modifying Aggregation Set Definitions You can modify an aggregation set definition before that definition has been sent to the host software If an aggregation set has been defined to the host you cannot modify the definition You can only display the aggregation set definition Note To modify an aggregation set after it has been defined to the host you can create a new aggregation set definition see 6 3 2 with the necessary attributes and then optionally delete the incorrect aggregation set see 6 3 4 To display an aggregation set or modify an aggregation set definition before it has been defined to the host select an aggregation set from the Aggregation Sets list box in the Aggregation Set Maintenance d
515. value for each parameter reference and inserts it into the function call MIN value This function compares all the values and returns the lowest of the values MAX value This function compares all the values and returns the highest of the values AVG value This function calculates and returns a simple average using the values passed as input SUM value This function returns the sum of all the values passed as input TQ 01011 18 8 3 Customizing Derived Statistics 8 3 Derived parameter values are calculated at report time This calculation is done using the aggregation set specified for the report The GRAN function can be used to obtain the granularity of the aggregation set being used in the report This function is called without any parameters GRAN This function returns the granularity in seconds of the aggregation set being used for the current report Maintaining Derived Statistics with TeamQuest Performance Software Administration To perform any administration tasks using TeamQuest Performance Software Administration you must have a session established to the host software see 4 2 You access the dialog boxes for performing administration tasks by selecting the Administration menu Note When derived statistics maintenance tasks are being performed the database is locked Only the user performing the tasks can access the database for modifying or updating derived statistics
516. ved statistic is actually more flexible when substitution is used however there may be times when you want to reference fully qualified values For more information on how to qualify an alias definition see 8 3 3 2 See 8 2 2 for more information on the qualifier substitution supported by TeamQuest Baseline 8 3 3 2 Qualifying Parameter References To qualify the statistic reference for an alias select the alias from the Alias List edit box of the Derived Statistic Definition dialog box and click the Qualify button The Qualifiers dialog box is displayed Qualifiers Primary Keys Category Group H STAT Log Category Workload Subcategory Total Statistic FCSS Read Func Regs Qualifier Keys System SYSTEM Workload Set swLs Workload Set swL 8 12 TQ 01011 18 8 3 3 3 8 3 4 Customizing Derived Statistics The Qualifiers dialog box contains a list of the primary keys Category Group Category Subcategory and Statistic for the statistic referenced in the alias It also contains list boxes for the qualifiers that can be used for the statistic Qualifier list boxes can appear for system workload set workload resource and application group The qualifier list boxes that are displayed depend on the statistic that is being qualified Only those qualifiers that have a relationship with the statistic are displayed The System Resource Workload Set Workload and Application Group fields contain drop down
517. very and audit control processing Percentage of Exec symbionts processing Percentage of miscellaneous switchable Exec processing Percentage of user real time processing A 11 OS 2200 Parameters A 12 ILES CMS 1100 UDS TIP Critical Deadline Demand Batch Category Group Category Resource Subcategory Statistic Name Hardware Default Unretryable Hwr Chk Diagnostic Interrupt 3 Interrupt 4 Interrupt 5 Interrupt 6 Interrupt 7 Reference Violation Addressing Exception Terminal Addr Exception RSC Underflow Overflow Signal Test and Set Invalid Instruction Page Exception Arithmetic Exception Data Exception Operation Trap Breakpoint Quantum Timer Interrupt 21 Interrupt 22 Interrupt 23 Percentage of independently linked Exec subsystems Percentage of CMS 1100 processing Percentage of UDS processing Percentage of TIP processing Percentage of critical deadline processing Percentage of demand processing Percentage of batch processing Processor Processor IPO IP1 Hardware Interrupts Hardware default interrupts per second Unretryable hardware check interrupts per second Diagnostic interrupts per second Interrupt 3 not used interrupts Interrupt 4 not used interrupts Interrupt 5 not used interrupts Interrupt 6 not used interrupts Interrupt 7 not used interrupts Reference violation interrupts per second Addressing interrupts Terminal addressing exception interru
518. w Statistic Fields dialog box is displayed This dialog box contains a complete list of all the key names that make up a statistic 8 14 TQ 01011 18 8 3 6 8 3 7 Customizing Derived Statistics Importing Derived Statistic Definitions To import all derived statistic definitions click the Import button on the Derived Statistic Maintenance dialog box see 8 3 1 When the Open File dialog box is displayed select the file that contains the derived statistic definitions that you want to import When the derived statistic definitions are imported all existing derived statistic definitions are updated to match the imported definition New derived statistic definitions are inserted into the Derived Statistic Maintenance dialog box If a syntax error occurs the derived statistic definitions are not inserted in the Derived Statistic Maintenance dialog box After an import function your updates can be deleted by clicking the Cancel button on the Derived Statistic Maintenance dialog box To accept the updates click OK and all updates are written to the session database See 8 4 1 for an example of a derived statistics definition file Exporting Derived Statistic Definitions To export all derived statistic definitions click the Export button on the Derived Statistic Maintenance dialog box see 8 3 1 When the Save File dialog box is displayed select or enter a file that will contain a derived statistic definition for all derived
519. w accesses data in the performance database and displays the data in graphic or tabular format E ER See Executive request Exec See Executive Executive Exec A software program supplied by Unisys that controls the system operating environment The Exec processes user runs controls files manages system resources and performs input output operations for the user The Exec is the central component of the operating system and enforces security control Executive request ER A set of programmed instructions that let the Exec perform service functions such as reading and printing data and terminating a program Selected ERs affect system security The ability to execute these ERs is controlled and this lets specified users override security restriction and validations G granularity The time interval between points of data in the performance database Glossary 2 TQ 01011 18 Glossary H history report One of the Team Quest Baseline and TeamQuest Online report types History reports show performance data that has been stored in the performance database They let you view past system performance and examine performance trends host software The part of TeamQuest Baseline or TeamQuest Online that resides on the OS 2200 system and is responsible for collecting performance data storing the data in the performance database and maintaining the database The host software is the server in the client server model of the Tea
520. w button indicating that you are going to insert a workload set definition into the database 4 Workload Maintenance Session cl01 Database PRODUCTION R Workload Info primary iotrace primary mprlog primary dia m workload set Import Export mot Emot Dee 4 Workload Set Definition Session cl01 Database PRODUCTION Source Data Type hyslog Workload Set Name Active Cancel Help TQ 01011 18 5 3 3 Customizing Workloads Workload Set Definition Dialog Box The Workload Set Definition dialog box contains information about the session database and source data type for which you are creating a workload set definition It also contains the following fields Workload Set Name supplies a name for the workload set A default name is supplied from TeamQuest Baseline You can use this name or supply your own name This field is limited to 51 characters in length Active specifies the status of the workload set A check in the active box indicates that the workload set is active and is available for receiving data from the data collection probes If this box is not checked the workload set is inactive and does not receive data When your definition is complete click OK The new workload set is inserted into the Workload Info list box in the Workloads Maintenance dialog box The workload set appears after the item selected in the Workload Info list box To define the wo
521. were available for this sample Processor Total Utilization Percent of time idle including forced idle Percent of tme in forced idle Minimum percent of time idle Maximum percent of time idle Percent of time in Exec processing Minimum percent of time in Exec processing Maximum percent of time in Exec processing Percent of time in user processing Minimum percent of time in user processing Maximum percent of time in user processing Percentage of miscellaneous nonswitchable Exec processing Percentage of Exec I O processing Percentage of Exec processor control processing Percentage of Exec memory management Percentage of Exec TIP integrated recovery and audit control processing Percentage of Exec symbionts processing Percentage of miscellaneous switchable Exec processing Percentage of user real time processing Percentage of independently linked Exec subsystems Percentage of CMS 1100 processing Percentage of UDS processing OS 2200 Parameters A 10 TIP TIP MIN TIP MAX Critical Deadline Demand Batch SCP Factor SCP Scaled Factor Performance Level Category Group Category Subcategory Statistic Name Hardware Default Unretryable Hwr Chk Diagnostic Interrupt 3 Interrupt 4 Interrupt 5 Interrupt 6 Interrupt 7 Reference Violation Addressing Exception Terminal Addr Exception RSC Underflow Overflow Signal Test and Set Invalid Instruction Page Exception Arithmetic Excepti
522. wledgement The host database server had acknowledged a data request sent to it but not within the sessions communications timeout time If the time listed in the Session Timeout window is past the current time on the host your communications session may have been destroyed due to a failure on the host You should close the session by selecting File gt Terminate Session If the host replies to the outstanding data request TeamQuest Performance Software Administration updates the Session Timeout window TeamQuest Performance Software Administration automatically updates the Current Timeout Value to reflect the amount of time the host took to respond to the message TeamQuest Performance Software Administration leaves this window open to let you know that communications on the session has slowed down TQ 01011 18 4 7 TeamQuest Performance Software Administration 4 7 Terminating the Communications Session To terminate a session 1 Select File gt Terminate Session The Administration Session Selection dialog box is displayed Administration Session Selection Name cl01 Socket cl01 2700 Host CLOT Host Type 2200 Cancel 2 From the list of open sessions select the session you want to terminate The session that you select is described in the session information on the right side of the dialog box 3 Click OK The selected session is terminated and you are returned to the main window When you have completed you
523. wn Subcoll Prb Valid Write Valid Dec Subcoll Cache Chg Owner Coll RLP Req Sec Switch Port Sec Init Dump Sec Purge Node Sec Down Subcoll Sec Prb Valid Sec Write Valid Sec Dec Subcoll Cache Sec Chg Owner Coll Sec TQ 01011 18 Total subcollection unlocks by UDS Total file unlocks by UDS Total string unlocks by UDS Total release block record unlocks by UDS Total record unlocks per second by UDS Total block unlocks per second by UDS Total process unlocks per second by UDS Total subcollection unlocks per second by UDS Total file unlocks per second by UDS Total string unlocks per second by UDS Total release block record unlocks per second by UDS IO Activity Multihost UDS RLP Misc Total number of UDS request to the RLP Total switch ports by UDS Total initial dumps by UDS Total purge nodes by UDS Total down subcollections by UDS Total probe validity checks by UDS Total write validity checks by UDS Total declare subcollection caches by UDS Total change owner collection changes by UDS Total number of UDS requests to the RLP per second by UDS Total switch ports per second by UDS Total initial dumps per second by UDS Total purge nodes per second by UDS Total down subcollections per second by UDS Total probe validity checks per second by UDS Total write validity checks per second by UDS Total declare subcollection caches per second by UDS Total change owner collection changes per second by UDS A 35
524. xecuting in SIP Time the processor was executing in TIP Time the processor was executing in real time Time the processor was executing in deadline batch Time the processor was executing in demand Time the processor was executing in batch Processor Processor IPO IP1 Util by Switching Level Time the processor was executing in Exec 0 Time the processor was executing in Exec 1 Time the processor was executing in Exec 3 Time the processor was executing in SIP Time the processor was executing in TIP Time the processor was executing in real time Time the processor was executing in deadline batch Time the processor was executing in demand Time the processor was executing in batch TQ 01011 18 OS 2200 Parameters A 2 2 1 Performance Monitor The following statistics are available on 2200 systems only Parameter Hierarchy Category Group Category Subcategory Statistic Name Available Time Available Category Group Category Subcategory Statistic Name Idle Forced Idle Idle MIN Idle MAX Total Exec Total Exec MIN Total Exec MAX Total User Total User MIN Total User MAX Misc NonSwitchable IO Processing Processor Control Memory Management TIP Integ Rec Aud Cntrl Symbionts Misc Switchable Real Time ILE CMS 1100 UDS TQ 01011 18 Processor Total Collection Base Total available time for all IP performance monitors for this sample Percent all IP performance monitors
525. xit gt gt Resource Name precede choice with to deselect gt 1 IP000 gt 2 IP001 gt 3 IP002 gt 4 IP003 gt gt Enter number name lt blank gt for more or E to exit gt 2 gt gt Resource Name precede choice with to deselect gt 1 IP000 gt 2 IP001 gt 3 IP002 gt 4 IP003 gt gt Enter number name lt blank gt for more or E to exit gt e E The following lettered descriptions correspond to the capital letters in the preceding example The letters are not part of the actual screen display A TQ 01011 18 Enter c at the prompt to create a new threshold definition The tqzip program then displays a series of prompts for you to select the name of the parameter or derived statistic for which you want to define thresholds The tqzip program displays a list of the system names in the database To define thresholds for derived statistics you must select DERIVED from this list You can enter either the number or the name DERIVED The tqzip program displays a list of the derived statistics in the database Select the derived statistic name for which you want to create a threshold definition You can enter either the number of the statistic or the statistic name 7 15 Customizing Alarms and Thresholds 7 9 7 16 A derived statistic may require qualifiers to calculate its values Qualifiers can be system names resource names workload set workload or application gro
526. xt is printed The value must be a positive integer Foreground Indicates the color in which the text of the annotation is printed Background Indicates the color of the annotation box that appears directly behind the annotation text If a background color is specified the annotation box covers any graph lines and other text that may appear behind this area If the Background field is not specified TeamQuest View allows any graph lines or other text that may be behind the title to show through Border Indicates the color of the border line that appears for the background area of the annotation text If the Border field is not specified TeamQuest View does not display a border line around the background area of the annotation text The default values for the text attributes are Arial Normal 8 and Black Export Data Page Spread Comma The specifications for exporting data from TeamQuest View Page Indicates whether the data from the current page of the report is exported The acceptable values are ON or OFF ON indicates that only the data from the current page of the report is exported OFF indicates that all data from the report is exported Spread Indicates whether the data is exported in spreadsheet format The acceptable values are ON or OFF ON indicates that the data is output in spreadsheet format OFF indicates that the data is output in a database like format C 26 TQ 01011 18 Report Definition File Comma I
527. y Statistic Name Reqs Sec Category Group Category Resource Subcategory Statistic Name Reqs Sec System Exec Requests 10 DATES Total Requests per second for each Executive Request ER System Exec CALLs CREATE GATE DELETE GATE Total Requests per second for each call gate Processor Statistics The Processor category group provides summary statistics for general processor Hardware Monitor and Performance Monitor utilization The following statistics are available on all systems with SIP level 4 or higher configured Parameter Hierarchy Category Group Category Subcategory Statistic Name Avg Queue Size Queue Searches Processor Total Switching Average number of activities waiting to be dispatched Number of times the switching queue was searched The following statistics are available only on 2200 systems with SIP level 5 configured Parameter Hierarchy Category Group Category Subcategory Statistic Name Exec 0 Exec 1 Exec 3 TQ 01011 18 Processor Total Util by Switching level Time the processor was executing in Exec 0 Time the processor was executing in Exec 1 Time the processor was executing in Exec 3 OS 2200 Parameters A 8 SIP TIP Real Time Deadline Demand Batch Category Group Category Resource Subcategory Statistic Name Exec 0 Exec 1 Exec 3 SIP TIP Real Time Deadline Demand Batch Time the processor was e
528. y performing the following steps 1 Using tqzip define the thresholds you want to use for your day shift see 7 4 2 Unload the current set of threshold definitions to a file called DAYTHRSHLDS see 7 13 3 Use FURPUR to copy the DAYTHRSHLDS to a new file called NITETHRSHLDS TQ 01011 18 Customizing Alarms and Thresholds 4 Edit NITETHRSHLDS to contain the threshold values you want to use during the night shift 5 Schedule two separate batch jobs to be run each day The first job will run at the start of the day shift It will execute tqzip with the l and t options to load the threshold definitions from the DAYTHRSHLDS file The second job will run at the start of the night shift It will load the threshold definitions from the NITETHRSHLDS file 7 15 Turning Threshold Checking On and Off The tqzip program allows you to control whether or not the TeamQuest Baseline probes perform threshold checking on the data they collect When threshold checking is turned off the probes do not check the parameter values for threshold conditions When threshold checking is turned on the parameter values collected by the probes are checked at each sample for threshold violations 7 15 1 Using the Alarms and Thresholds Menu To turn off or on threshold checking select Turn Threshold Checking On Off from the Alarms and Thresholds Menu see 7 3 by entering 5 at the prompt If threshold checking is currently turned on for the performanc
529. y planning and what if tuning scenarios Typical users include performance analysts and capacity planners Introduction 1 3 What Is TeamQuest Baseline TeamQuest Baseline is the OS 2200 software product that provides the services for the TeamQuest Performance Software family on your OS 2200 system The TeamQuest Baseline must be installed on every monitored host and consists of the components that collect store manage and administer the performance data Team Quest Baseline performs the following functions for the TeamQuest Performance Software family Data collection and storage Data is collected by probes and stored in a performance database You can select from a predefined set of probes or you can define new probes Administration You can control the data that is collected how often it is collected and how it is stored You can also derive statistics and characterize workloads Alarms and Thresholds You can specify conditions and actions for the TeamQuest Baseline to test and act upon Application interfaces TeamQuest Baseline makes the collected and stored data available to the performance applications TeamQuest View and TeamQuest Model using secure application interfaces TQ 01011 18 Introduction 1 4 What Can You Do with TeamQuest Baseline and TeamQuest Online The TeamQuest Baseline is the required product within the TeamQuest Performance Software family TeamQuest Online is a subset of TeamQuest
530. ystem control TMSC processor started The current number of active threads The maximum number of active threads since the transaction management system control TMSC processor started The current number of threads reserved for servers that are not active The maximum number of configured threads Server Table Information The following information is stored in the performance database tables Table Field Hierarchy Class 2200 Name File name odtpfile Collection interval Default retention 1 day Data Item Type Time Double Interval Double Server String Type String ReqQ Double Fails Double Aborts Double Sleeps Double Sleep_Time Double Receives Double Sends Double ODTP Server Based on the table collection interval Description The time this information was recorded in the database for this sample The desired sample interval in this collection period The name of the server The server type TIP Batch TIP Sched The number of requests queued in this sample interval The number of failures in this sample interval The number of aborts in this sample interval The number of sleeps in this sample interval The sleep time in this sample interval The number of receives in this sample interval The number of sends in this sample interval TQ 01011 18 OS 2200 Parameters N2PC_Calls N2PC_Regs 2PC_Calls 2PC_Reqs Count 2PC_Txns 2PC_Time Txn Txn_Commits Txn_Rollbacks Txn_Timeouts TQ 0
531. zip d databasename o parm file TQ 01011 18 where Customizing Performance Databases specifies the database that tqzip uses The databasename parameter is the name of a database that has been created on the host system If the d option is not specified the Database setting in the tq ini element is used see 3 1 outputs the database parameter list to the file specified by the parm file parameter Parameter File Format The o option of the tqzip program generates the list of parameters and places the list in the file you specified as the parm file parameter on the processor call If you edit this file to delete parameters from the database see 10 9 you should maintain the file in the format shown The following shows how the list of parameters appears in the file Sys BEGIN PARAMETER LIST created from lt PRODUCTION gt on Fri Jan 13 14 05 45 2012 A tem CatGroup Cat SubCat Stat Res App WL Set WL System Elapsed Time B Processor Processor Utilization Real Time IP000 Processor Processor Utilization Real Time IPOO1 Processor Processor Utilization Real Time IP002 Processor Processor Utilization Real Time IP003 Processor Processor Utilization Exec0 IP000 Processor Processor Utilization Exec0O IPO01 Processor Processor Utilization Exec0O IP002 Processor Processor Utilization Exec0 IP003 System Throughput TIP Txns Sec System Thr

Download Pdf Manuals

Related Search

Related Contents

据付説明  PROFESSIONAL WIRELESS INTERNET WEATHER STATION    Xcellon RAID System Software User Manual.indd  FINAL USER 4 BANKERS REALM  GUIDA PER LE CONNESSIONI ALLA RETE  MBkeyprog-2  375 Field Communicator  JTAE471 取扱説明書  SES E 7210-4  

Copyright © All rights reserved. Failed to retrieve file