Home

HP GlancePlus User`s Manual

1. Contents 2 CPU Bar Disc Bar Summary Fields Filter Processes Workload Screen Interesting Applications Interesting Application Fields Top CPU and Disc Consumer CPU Detail Screen Type of Utilization Queue Distribution Miscellaneous Performance Metrics Top CPU Consumer Disc Detail Screen Individual Disc Utilization Summary Disc Utilization Top Disc Consumer Memory Detail Screen Fault Types Miscellaneous The Expert Facility Invoking The Expert Facility What Does The Expert Facility Do Recommendations For Using The Expert Facility Changing The Expert Facility s Rules Customizing HP GlanceP lus Customizing Startup Refresh Time Selecting the Initial Screen Customizing GlancePlus Global Bars Global CPU Bar Lo Global Memory Bar Customizing Configuration Interesting Process Thresholds Display Control Parameters The Expert Parameters 5 30 5 30 5 31 5 32 5 34 5 34 5 35 5 36 5 37 5 38 5 39 5 39 5 43 5 44 5 45 5 46 5 46 5 47 5 48 5 49 6 3 6 3 6 7 6 8 7 1 7 1 7 1 7 2 7 2 1 3 7 4 1 7 7 11 7 13 Contents 3 Additional Parameters o 7 14 GlancePlus Configuration File Security 15 Configuring the Expert Facility 7 15 Highlighting Metrics a FLG Performance Recommendations o ool wd A Multiprocessor Systems B Wait Reasons Expanded Definitions C Detailed WAIT Reasons D Application Definition Files Application D
2. Press RETURN to continue then press X for more details or 0 for an overview summary Press X again Low priority processes D E gt are being starved for CPU If you want the lower priority processes to run consider lowering the priority of the CPU consumers listed below Action QZAP pin 122 PASXL gt for MEL EELKEMA from C to D queue CPU is heavily used by high priority processes C queue and above The following processes are using a large amount of CPU time m JSNO Dev Logon Pin Program Pri CPU Disc Trn Resp Wait 1009 46 MEL EELKEMA 122 PASXL C 91 9 0 8 O 0 08 87 The following processes were starved for CPU time m JSNO Dev Logon Pin Program Pri CPU Disc Trn Resp Wait JZ435 10 PAYROLL ACCTNG 14 PRNTCHEK D 0 0 8 8 O 0 8 100 Even if no action is recommended it can be helpful to display the high level analysis The analysis might indicate why no action should be taken at this time You might discover that even though your system CPU is 100 busy due to interactive users and a CPU bottleneck is indicated no action is recommended since no process is starving for the CPU Continued use of the expert high level analysis can teach you some of the subtleties of managing performance You may not always agree with the suggested actions but the information can help you understand your system better There is one level of analysis available beyond the high level discussed above If after viewing the h
3. Since OFICPROD is included in Order Processing any programs in FURNITUR OFICPROD or ELECTRON OFICPROD will be logged in the Order Processing application not in Office Products However if the order of the applications is reversed any programs in the groups FURNITUR and ELECTRON in the OFICPROD account will be logged in the Office Products application not in Order Processing By default no user applications are defined Note Since any process on the system can belong to only one application no process will be counted in more than one application The application definition file is processed in the order entered and the first match of a program name user logon and queue will define the application to which a particular process belongs FILE The FILE filename parameter specifies which program files belong to an application It includes all interactive or batch executions of these programs and applies to the last APPLICATION directive issued An error is generated if no APPLICATION directive is found The filename can be any of the following a A simple fully qualified MPE file name For example EDITOR PUB SYS m A generic file name For example VOODO0 a A partially qualified file name For example SPOOKS In this case all groups and all accounts are assumed SPOOK5 Multiple file names can be entered on the same line separated by commas or in separate FILE parameters D 4 Application Defin
4. Reads R lt 48 5 Writes oo oo o o lt 11 9 Memory Manager gt Reads ooooooo lt 1 1 Writes oo oo o o lt 1 9 Top DISC Consumer is Pin 306 at 2 1 0 s per sec LOGCHK PUB SECURE Figure 5 22 Disc Detail Screen Summary and Top Disc Consumer 5 46 Screen Metrics and Statistics Memory Detail Screen The Memory Detail screen displays metrics that show physical memory activity by the system during the current interval The detailed data area of the display is divided into two parts m Fault Types a Miscellaneous HPB1787 B 4 HP GlancePlus 13 86 28 Memory Detail O O O O O O O current avg high CPU HHS sE p BE B 587 50 76 Disc M MS SI x IB Bolle ee eee eee i 62 45 101 Mem amp ME ES SD DE 3 100 100 1007 Transient E lt 14 7 sec Permanent o rrr cece cree cece ee ener ee eee eee eee eee eee ee eee eee eee eees lt 0 sec NM Stack N lt 4 6 sec CM Stack E Le ce cc ee ee ee eee ee eee ee eee eee eee ee eee eens lt 2 4 sec NM Oo ea lt 0 sec CM Code coccoccccocccccoocccooo rr lt 0 sec CM Data E lt 1 8 sec File Object E EjJ o oooooooooooonooonononnonrrnnannnnonornrorooo lt 1 6 sec NM Sys Lib oooooooooooooonornonnrrnrrnrnoao nooo oo ooosrrrrooo lt 0 sec CM Sys Lib Loc ee cc ccc cc ce eee ee eee ee eee eee eee eee eee eens lt 0 1 sec MemMgr Fe ec ce nooo ooo ooo lt 47 Page Fault Rate E A lt 25 2 sec P
5. 1 0 The process is waiting to use the central processing unit On busy systems higher priority processes can consume all of the available CPU resources while lower priority processes must wait The PRI column displays the queue and priority number of a process On extremely busy systems a low priority batch job high PRI number can remain in a CPU wait state while all available CPU resources are consumed by higher priority processes lower PRI number When SQL applications are running with multiple users the database monitor process DBMON will stay in the DBMS wait state during periods of inactivity This wait state is unique to MPE iX SQL applications When TurboIMAGE applications request a lock the process will be semaphore waited SEM The process is waiting for the completion of a read to a file or write to a file This does not include memory manager I O which is accounted for in the MEM wait category On MPE iX systems main memory is used to buffer I O requests this greatly reduces the amount of time a process waits on DISC The son process is waiting for its father to reawaken it When one process creates a second process the first process is referred to as the father and the second process as the son The father process has the capability of starting and stopping the son process The process is impeded This wait state is used for a variety of reasons On MPE iX systems processes using databases make use of the D
6. 2 2 0 3 15 4 Commands Additional Information Overview o 4 1 Selecting Jobs Sessions and Processes gt J S P 4 o 4 1 MPE Command lO 4 2 Refresh Screen R 0 4 2 All Processes A 2 2 e o 4 2 Change Queue Q 4 3 List Users of a File U 4 3 Zero Averages Z o 4 4 Previous Display lt 2 2 4 4 Next Logical Display gt e 4 4 Ninety second Timeout on Command Prompts a 4 4 5 Screen Metrics and Statistics Global Screen e 5 2 Interesting Processes e 5 2 Interesting Process Components a 5 4 Job Session Screen Le 5 7 Job Session Bars and Summary Fields Lo e 5 8 Interesting Processes A Interesting Process Components a BLT Process Screen ee ee ee eee 13 Process Bars e DlA Process Summary Fields a 5 5 Process State Bars 16 STATE Components of Response Time 16 LAST Components of Response Time 5 17 Miscellaneous Process Information 2 2 2 2 4517 Process Screen Function Keys 2 5 19 Setting Option Values V 2 2 2 ee ee ee 5 19 Simple File Display 2 2 2 ee eee 5 21 Detailed File Display 2 2 2 2 2 2 2 5 21 Filter Selection Menu Variables 2 2 1 5 26 Filter Display Screen 2 2 2 ee ee 5 29 Filter Bars a ee 30
7. 3 subcommand is valid only during the Process screen display The first time you use the Display Trace key a 5 to 30 second delay occurs This is normal Note however that a user will not be able to execute this command without System Manager SM capability Consult the MPE XL System Debug Reference Manual for detailed information on how to read a trace 5 24 Screen Metrics and Statistics SPT Collect This function key allows you to use the HP SPT Software Performance Tuner product to create and run a program under HP GlancePlus for collecting metrics on the process currently running To display this process without using function keys press 4 The 4 subcommand is valid only during the Process screen display This command is optional and will operate only if the HP SPT product was installed The data SPT collects about a process is placed in the default log file SPTLOG The default time for collecting process metrics via SPT is 5 minutes but can be changed by using the job control word GLANCESPTTIME Collection times between 1 and 60 minutes can be selected when using SPT from within GlancePlus The file SPTOUT is used for STDLIST messages created from the SPT process running under HP GlancePlus If desired you can terminate a SPT process created by GlancePlus by using the MPE command and issuing the command SETVAR SPT_STOP TRUE The existence of the SPT_STOP variable will cause the SPT process to terminat
8. Group Access A user must have read lock and execute access to the GlancePlus files before running it Since GlancePlus is usually installed in PUB SYS all users typically would have the required access Installing the Program Usually HP GlancePlus is installed automatically through the MPE X AUTOINST process To customize the use of GlancePlus activate the GlancePlus UDC user defined command file as follows m Move the user defined command GLANCE from the GLANUDCX file to the desired UDC file and activate it or m Activate the HP GlancePlus UDC file and type the following SETCATALOG GLANUDCX 2 2 Installation Getting Started Overview This chapter shows you how easy it is to use the HP GlancePlus program You will learn how to run the program display a screen move to other screens get help print screens exit the program and use the single letter commands to invoke GlancePlus s various tasks Running HP GlancePlus for MPE iX To execute the program a Type GLANCEXL PUB SYS m If you have activated the HP GlancePlus UDC file type GLANCE When you become familiar with GlancePlus you may want to customize the GlancePlus startup procedure Chapter 7 contains information on customizing the GlancePlus startup procedure Getting Started 3 1 Exiting HP GlancePlus To exit the program at any time m Press E or Fs from any screen except an online help screen More Than a Loo
9. NBRATES PUB ACCTG KSAM 255 1737 PERM IN SHR L 3 NBRKEYS PUB ACCTG KSAMK 4z 1282 PERM IN SHR 3 AZB4K cc eee ee eee eee 2338 AUDIT LOG ACCTG 32529 32529 PERM APEN SHR 3 VERRMSGS PUB SYS 1414 1766 PERM IN 35 Press RETURN to continue Cor enter a GLANCE command Display Display Display SPT Other Refresh Help Exit Files Family Trace Collect Keys Screen Glance Figure 5 11 File Display Screen Detailed Display Screen Metrics and Statistics 5 21 where File Group Account is the formal designator for the file and File The file name or if unnamed a question mark 7 Group The group name or if unnamed a series of dots Account The account name or if unnamed a series of dots Type File code mnemonic as in LISTF 1 command or File code number user defined file code or One of the following specially defined mnemonics RXRAW LaserRX raw log file file code 1135 RXEXT LaserRX extracted log file file code 1136 ROOT TurboIMAGE data base root file file code 400 DSET TurboIMAGE data base data set file code 401 DBREM TurboIMAGE remote data base file code 402 DBILR TurboIMAGE ILR file file code 404 PRIV Any other negative file code or if the file code is zero KSAM Foptions file type 1 RIO Foptions file type 2 KSAMX Foptions file type 3 CIR Foptions file type 4 SPOOL Foptions file type 5 MSG Foptions file type 6 Rec Pointer Th
10. Values Job Control GLANCNFG File Default Command Word Keyword Value FILEDISP GLANCEFILEDISP FILEDETAIL DETAIL PROCKEYS GLANCEPROCKEYS PROCKEYS DEMAND DISPCPU GLANCEDISPCPU DISPLAY CPU INTER BATCH DISPMEMORY GLANCEDISPMEM DISPLAY MEMORY CONTENTS n a GLANCECLEARSCREEN CLEARSCREEN ON OFF OFF Expert Facility Values Job Control GLANCNFG File Default Command Word Keyword Value XPERT GLANCEXPERT XPERT DEMAND XLEVEL GLANCEXLEVEL XLEVEL LOW Additional Parameters Values Job Control GLANCNFG File Default Command Word Keyword Value n a n a XPERT ANY n a n a QZAP SM n a n a QZAPSYS NONE 1 T 5 GLANCETIME REFRESH 30 Seconds 7 6 Customizing HP GlancePlus Interesting Process Thresholds Interesting process thresholds set on the Global and Filter screens allow you to select processes to be displayed and fields to be highlighted The MINIMUM CPU parameter determines the minimum amount of CPU time a process must consume in order to be displayed DEFAULT 1 percent GLANCNFG THRESHOLD MINCPU percent JCW SETJCW GLANCEMINCPU percent VALUES MinCPU percent The CPU parameter determines how much CPU time a process must consume to be displayed with its CPU field highlighted DEFAULT 10 percent GLANCNFG THRESHOLD CPU percent JCW SETJCW GLANCECPU percent VALUES CPU percent The DISC parameter determines what disc IO rate a process must perform in order to be displayed with it s DISC field highlighted DEFAULT 10 I0s
11. With XLEVEL LOW and XPERT DEMAND the Expert facility adds no additional overhead Press X to request an on demand global analysis at any time Since the overhead of the Expert facility on most systems is negligible you can safely let it run in continuous mode You may use XLEVEL LOW and XPERT CONTINUOUS to report any global problems and to accumulate the overview summary information for the 0 command Using XLEVEL HIGH should not add significant overhead on most systems You can see the overhead by using GlancePlus to monitor itself under various settings on your system Hint The frequency of screen updates is one factor that affects GlancePlus s overhead most directly If you set the update interval to its lowest setting 5 seconds or press Return or repeatedly to generate rapid updates the CPU overhead can become very high If you want to collect long term summary information set a longer update interval 60 90 seconds to lower the GlancePlus overhead substantially However in some cases setting the update interval to longer than 90 seconds might not reduce the GlancePlus overhead To keep process information from backing up in the measurement interface information should be extracted every 60 90 seconds Failure to do so can result in missing and inaccurate data To avoid this problem if you select an update interval longer than 90 seconds GlancePlus extracts the process data automatically once a minute to pre
12. banner line 3 3 BATCH parameter D 5 Cc capabilities group 2 2 CCTL 3 15 changeable parameters 7 5 Change Queue command 3 10 4 3 changing priorities 4 3 clear screen parameter 7 12 clock cycles 7 16 rate 5 50 color emphasized metrics 5 3 6 2 7 16 Command Interpreter 5 4 5 10 5 11 5 32 5 36 5 43 Command List key 3 9 commands 4 1 4 aaa key 3 10 4 1 Index 1 aaa key 3 10 3 15 4 2 5 4 5 11 5 25 aaa lt key 3 10 4 4 aaa gt key 3 10 4 1 4 4 or 3 13 aa A key 3 10 4 2 aa C key 3 9 3 10 5 37 aa D key 3 9 3 10 aa E key 3 2 3 10 eh key 3 10 5 26 aa G key 3 10 aa H key 1 3 3 9 3 10 3 13 6 2 aa 1 key 3 10 aa J key 3 10 41 aa L key 3 10 oe key 3 10 aa O key 3 10 6 4 aa P key 3 10 3 14 4 1 aa Q key 3 10 4 3 aa R key 3 10 4 2 eS key 3 10 4 1 aa T key 3 10 aa U key 3 10 4 3 aa v key 3 10 3 16 5 3 5 19 5 20 6 5 6 7 7 4 aa w key 3 10 aa x key 3 10 3 12 6 3 6 5 aa Z key 3 7 3 10 4 4 a ctri Q key 3 10 a Ctrl S key 3 10 al Ctrl Y key 3 10 al Enter key 3 10 al Expert Analysis key 3 12 All Processes 3 10 4 2 al Space Bar key 3 10 available list 3 9 Change Queue 3 10 4 3 CPU Detail 3 9 3 10 5 37 Disc Detail 3 9 3 10 Exit Glance 3 2 3 10 Expert 6 3 Index 2 Expert Analysis 3 10 3 12 6 5 Expert Overview 3 10 6 4 Filter 3 10 5 2
13. the screen identifier The detailed data area of the display is divided into five parts a Process Bars m Process Summary Fields m Process States m Miscellaneous Process Information m Process Screen Function Keys HPB1787 B 04 00 HP GlancePlus 11 25 58 Process Pin 198 J515 Ma current avg high CPU MSST IB Be Pl 90 897 100 Disc MS SI E i z5 28 59 Mem M MC ES SD DE 100 100 1007 CPU E eee lt EERE CMPRG TAS3Q9 P TASPRG MM SREL Disc Ki e cee cece ec cece ec ee cence eens lt 2 Elapsed CPU Tot Disc Tot 115 5 min 7 1 min 3 2 k State E ED DQ GP PQ i 1007 Last E CP PO 0 lt 1007 Wait READY QUEUE 34 Compatibility Mode 7 E lt 19 Queue Priority D224 Native Mode Switches N N lt sec Scheduling State Ready Comp Mode Switches Transactions min B Response Time Display Display Display SPT Other Refresh Help Exit Files Family Trace Collect Keys Screen Glance Figure 5 5 Process Screen Screen Metrics and Statistics 5 13 Process Bars The Process Bars show how the selected process is using CPU and disc resources CPU B Disc X Figure 5 6 Process Screen Process Bars The Process bar has the following components CPU DISC The CPU bar shows the percentage of central processing unit time used by the process during the last interval The codes on this bar correspond to the codes used in the Global CPU bar see CPU Bars under Readin
14. 04 00 HP GlancePlus 14 44 46 Global Ma current avg high CPU Ms SI X ERETTE 87 S4 J17 Disc MSL Mee cece cc cee cee ee cette tees eee eeeee i 12 9 31 Mem M MC ES SD DE 100 100 1007 JSNo Dev Logon Pin Program Pri CPUZ Disc Trn Resp Wait IXPERT e S GR S203 134 SHARON EVAL 182 MNTMENU 2 827 0 0 E 0 2 TERM S211 310 BECKY EVAL 81 MNTMENU 1 87 0 0 Mi 0 4 TERM S5209 23 COLLECT EVAL 212 REVALINQ 3 57 0 6 El 0 4 TERM SZ30 314 MARYANN EVAL 291 MNTMENU 2 124 0 1 E 0 3 TERM S5288 215 VIRGINIA EVAL 228 MNTMENU 2 124 0 1 i 0 1 TERM S208 227 VAN ACCTG 288 APPRB1 0 37 0 0 Z Y TERM SZ93 332 JOE TELESUP Z240 GLANCEXL 3 67 0 0 4 B 6 5295 334 DUP 13 DIALUP 312 CI 0 77 0 2 2 0 0 5295 334 DUP 13 DIALUP 237 LOGON 1 87 0 0 Z 2 5 5295 334 DUP 13 DIALUP 48 RESMNT 9 67 0 5 Z 0 4 5295 334 DUP 13 DIALUP 103 REVALINQ 0 97 0 2 4 0 0 SZ97 333 DUP 41 DIALUP 239 TTSSON 6 37 0 2 4 CPU 5298 225 HARBISON EVAL 144x 2 827 Q 1 2 0 0 SON 5298 225 HARBISON EVAL 261 5 67 0 7 2 0 0 TERM Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 3 6 An Example of the Expert Facility 3 12 Getting Started Help Facility H You can access the GlancePlus Help facility in three ways Enter the H command character press the Help softkey or enter a question mark A prompt will ask if you want to see a list of available commands access help specific to the current screen or access genera
15. 0500 Service time for an individual disc seconds MEM_CPU 8 CPU usage managing main memory percentage of total MEM_PFAULTS 30 Memory page faults faults second MEM_CLOCK 25 Memory clock cycles MEM_LIBFAULT 2 Memory page faults on library code faults second PROC_CPUSEC 3600 Time of total CPU usage by a process seconds PROC_CM 80 CPU usage in compatibility mode by a process percentage of total PROC_SWCM 100 Switches from compatibility mode to native mode by a process switches second PROC_SWNM 100 Switches from native mode to compatibility mode by a process switches second 7 16 Customizing HP GlancePlus Table 7 2 GlancePlus Highlighting Characteristics continued Hitem Default Description APPL_CPU 20 CPU usage by an application percentage of total APPL_DISC 50 Disk I O rate by an application IOs second APPL_TRANS 120 Transaction rate by an application transactions per minute Performance Recommendations The second area in which the Expert facility can be customized consists of the rules by which Expert facility accesses a system s overall performance also known as low level analysis The Expert facility defines a set of symptoms wherein each symptom represents a particular performance situation A set of rules define the conditions that increase the probability that a particular symptom is true Computer scientists might recognize the use of fuzzy logic a technique in which a condition does
16. 103 REVALINQ 0 97 0 2 4 0 0 SZ97 333 DUP 41 DIALUP 239 TTSSON 6 37 0 2 4 CPU 5298 225 HARBISON EVAL 144x 2 827 Q 1 2 0 0 SON 5298 225 HARBISON EVAL 261 5 67 0 7 2 0 0 TERM Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 1 Global Screen Interesting Processes Only processes that you define as interesting are displayed on the Global screen This allows you to concentrate on processes most likely to affect performance A process can be interesting if it exceeds a user defined threshold value it was just created or it was terminated during the last interval 5 2 Screen Metrics and Statistics Interesting processes are scrolled down as new entries are added This keeps the latest information at the top of the list below the Global bars You can scroll the terminal display to examine previous processes and intervals again A blank line separates each group of process information from the group that was displayed during the previous time interval JSNo Dev Logon Pin Program Resp Wait XPERT CPU 1S HIGH S203 134 SHARON EVAL 18Z MNTMENU 0 2 TERM S211 310 BECKY EVAL 81 MNTMENU 0 4 TERM S5209 23 COLLECT EVAL 212 REVALINQ 0 4 TERM SZ30 314 MARYANN EVAL 291 MNTMENU 0 3 TERM S5288 215 VIRGINIA EVAL 228 MNTMENU 0 1 TERM S208 227 VAN ACCTG 288 APPRB1 Y TERM SZ93 332 JOE TELESUP Z240 GLANCEXL B 6 5295 334 DUP 13 DIALUP 312 CI 0 0 5295 334 DUP 13 DIALUP 237 LOGON 2 5 5295 334 DUP 1
17. Additional Information 4 3 Zero Averages Z Use this command to reset the averages for CPU Disc and Memory to zero and recompute them beginning with the current interval values These averages are shown on the screen in the column labeled avg and as an asterisk in the horizontal bars Previous Display lt Use this command to return immediately to the display that was last active before you entered the current display For example if you examine a job on the Job screen then move to the Process screen for an in depth look at a specific process you can use the Previous Display lt command to return to the Job screen Next Logical Display gt Use this command to move to the next logical display based on your current display As discussed earlier in this chapter the Next Display command provides a guided analysis capability you can use to zoom from the Global CPU Detail Disc Detail or Filter screen to the Process Detail screen of the top consumer of resources You can also use it to zoom from the Workload screen to the Filter screen to examine the group of application programs that consumed the most CPU during the last interval The Next Display command is only valid when issued while one of the screens mentioned above is the current screen Ninety second Timeout on Command Prompts Some commands issue a prompt and wait for your response Examples are the Set Option Values command and the MPE command A ninety second
18. GlancePlus 12 47 CPU MMS SI IB Disc M MS SI Mem M MC ES SD Type Utilization MeEMMgr nc cee ce eee eee ee ee eee lt Y System S ewes lt 1 Session E Tose cee ee ee eee lt 34 ICS ke eee cee eee eee ween lt 7 Queue E BE Switches To CM 190 sec Switches To NM S lt 1965 sec Interval CM7 lt 237 Interval Process Completions 13 sec Top CPU Consumer is Pin 1286 at 39 8 Global Select Print MPE Command aa current avg high 36 CPU Detail x BE Bl 75 64 76 x IB OB ee i 86 68 181 DE 100 100 1007 Type Utilization Dispatch oooooooooooooooo o lt 17 Batch B Bocce eee eee ee eee lt 2 Pause P Powe cee ee lt 577 Idle ooooooooooom oo oo o o oo lt ar AS BS 26 CS 41 DS 33 ES 07 Current Ready Queue B lt Z Maximum Ready Queue D D lt 21 Launch Rate 131 sec CQ SAQ 19 Interval File Open Rate 26 6 sec Refresh Screen Exit Glance Time Adjust Help Figure 5 15 CPU Detail Screen Screen Metrics and Statistics 5 37 Type of Utilization This uppermost section of the detail area lists the percentage of total time used for different system resources and system purposes Utilization Type Utilization Loro eens lt Z Dispatch oooooooooooooooo 1 see eee eee eee weenie lt 14 Batch B IIS 2 rr lt 34 Pause P Poe ee eee lt 577 see ee eee eee eee eens lt Idle Lor ar Figure 5 16 CPU Detail Screen
19. HP GlancePlus is only one member of a family of performance products offered by Hewlett Packard If you need to go beyond immediate performance monitoring you may wish to consider other members of this family HP GlancePlus Pak for MPE iX combines GlancePlus with the ability to log data for analysis by programs on the HP 3000 or by spreadsheet packages on a personal computer a HP LaserRX MPE logs data similar to HP GlancePlus Pak for MPE iX but also contains a powerful performance data analysis tool It organizes and categorizes raw performance data into easy to use Microsoft Windows based graphs and charts that can be displayed on a PC You use a highly efficient hierarchical approach to analysis looking at data from the global to detail level a HP RXForecast MPE forecasts future resource utilization based on actual past history This tool helps you plan future resource requirements and can help you avoid periods of insufficient resources It requires HP LaserRX MPE as a companion program m HP SPT XL helps you optimize native mode programs 1 2 Introduction How to Use This Manual This manual will help you get maximum value from HP GlancePlus Depending on your needs and previous experience you can consult the following chapters for specific information m Chapter 2 for information on installing HP GlancePlus m Chapter 3 for information on the general features of the program m Chapter 4 for additional information
20. OP capability may erecute the QZAP command but not into or out of the B or linear queues QZAP NONE No user may execute the QZAP command The QZAPSYS parameter is only allowed in the GLANCNFG file It controls which users are allowed to execute the QZAP command against system processes Combinations of OP and SM are allowed in this parameter Care must be taken when adjusting system process priorities since this might degrade performance cause errors or even hang or crash the system DEFAULT NONE No user may QZAP a system process GLANCNFG QZAPSYS SM and or OP To allow users with SM and or OP caps 7 14 Customizing HP GlancePlus The REFRESH TIME parameter determines how many seconds between automatic updates of the GlancePlus display DEFAULT 30 seconds GLANCNFG REFRESH seconds JCW SETJCW GLANCETIME seconds VALUES Not used Use the I Interval or T Time command or press 15 GlancePlus Configuration File Security The GlancePlus configuration file enables certain capabilities and contains vital global configuration options Because of this a user is not allowed to use file equations to utilize an alternate configuration file If GlancePlus is installed in the PUB SYS group and the default access is still in effect any user who can log on in the PUB SYS group has the capability to change the configuration file You may want to restrict write and append access to MANAGER SYS via the ALTSEC command to furt
21. Process screen display The Display Files function appears in two formats depending on how the File parameter is configured File simple or File detailed This parameter is set using the Set Option Values V command The following formats occur 5 20 Screen Metrics and Statistics Simple File Display If the File parameter is configured File simple the Simple File Display format occurs where Filename rrr The file name comprised of the first eight characters of the formal designator for the file This field following the file name indicates the progress of the record pointer as compared with the end of file pointer A sample display might be DBO1 32 DBO2 45 Detailed File Display If the File parameter is configured File detailed the Detailed File Display format occurs as shown in figure 5 11 Open Files for Process Z1Z are File Group Account Type Rec Ptr EOF Domain Access Share Opens APPRO1 OBJ3 ACCTG NMPRG 659 789 PERM LOAD M SEM 3 AMRXL PUB ACCTG NMXL 8821 8937 PERM LOAD M SEM 13 UPDATES PUB ACCTG 447 447 PERM APEN SHR 1 APPRSCRN PUB ACCTG UFAST 879 1705 PERM IN SHR 3 APPRMAST 1 ce ee eee eee 92 APPRMAST PUB ACCTG KSAM 163598 1655088 PERM UPDT SHR L 52 APPRMASK PUB ACCTG KSAMK 18 148226 PERM UPDT SHR 52 NBMASTER oooooooooo 96 NBMASTER PUB ACCTG KSAM 988 3645 PERM IN SHR L 52 NBMASTEK PUB ACCTG KSAMK 74 Z770 PERM IN SHR 52 NBRATES oooooocooooo B
22. Processes HPB1787 B 4 HP GlancePlus 16 22 47 Session 551 O O O O O O O current avg high CPU ff AI IB E II 48 57 7 89 Disc E eee ccc ee cece ete eee e eee e ee eee eeeeees i 7 11 29 Mem amp ME ES SD DE 3 100 100 1007 CPU Xuuuusesusnesssossesosseeeeseo lt 1 2 4 DIANNE EVAL I 27 i DiSC X i ococoocooooooomoo roo roo ooooo lt 5 Elapsed CPU Tot Disc Tot i 8 1 hr i 1 5 min 1 3 k Pin Program Pri CPUZ Disc Trn Resp Wait Elapsed CPU Tot Disc Tot P Pin 144 REVALINQ C152 1 2 0 5 g 1 6 TERM 16 1 s 1 1 s 15 0 303 Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 3 Job Session Screen Screen Metrics and Statistics 5 7 Job Session Bars and Summary Fields A lt 1 2 lt 4 DIANNE EVAL see eee eee eee eee eee eee ee eee lt 5 Elapsed CPU Tot Disc Tot i 8 1 hr i 1 5 min 1 3 k The Job Session Bars appear below the Global bars CPU Disc and Memory and show additional information about how the selected job or session is using CPU and Disc resources The Job Session Bars and Summary Fields contain the following information CPU DISC This CPU bar shows the percentage of central processing unit time used by the job or session during the last interval The codes in this bar correspond to the codes used in the Global CPU bar see CPU Bars under Reading the Screen Display in chapter 3 The number at the right of the bar i
23. Type of Utilization These system resources and system purposes include the following MemMgr System Session Ics Dispatch Batch This line shows the time during the current interval that the CPU was processing certain Memory Manager activities such as fetches and swap ins If memory manager CPU utilization is consistently higher than 8 the system might be experiencing a main memory bottleneck This line shows the time during the current interval that the CPU spent on system processes such as NS services spoolers and the Loader The Session line shows the time the CPU spent on interactive processes during the current interval If more than 70 of the CPU time was spent consistently on interactive processes the system might be experiencing a CPU bottleneck condition The Interrupt Control Stack ICS line shows the time spent on interrupt activity during the current interval The ICS value is derived by subtracting the sum of all other CPU states during the interval from 100 This line shows the time the CPU spent on MPE iX Dispatcher activity This line shows the time the CPU spent processing batch jobs during the current interval 5 38 Screen Metrics and Statistics Pause This line shows the time the CPU was paused for the disc during the current interval If more than 25 of the time is spent consistently in a paused condition the system might be experiencing a disc bottleneck condition Id
24. and summary fields The top CPU consumer is identified by its PIN process identification number the percentage of CPU it used during the last interval and the program name or MPE iX Command Interpreter CI command last issued If an MPE iX system process is the top consumer it is identified by the program name or by the text running program on behalf of MPE ix The top Disc consumer is identified in the same way as the top CPU consumer except that the rate of disc transfers transfers second for the process during the last interval is displayed instead of CPU percentage If no disc activity occurred during the interval the Top Disc Consumer line will not appear on the screen Note If you do not have System Manager SM capability you will see the letters CI instead of the command string text when the program is a Command Interpreter 5 36 Screen Metrics and Statistics CPU Detail Screen If you choose the C command for the CPU Detail screen you can view metrics that are specific to CPU usage In addition to the Program Banner Line and Global Bars this screen will show CPU usage by the dispatcher memory manager system and batch processes as well as the time spent in various states such as paused idle or in a queue The detailed data section of the display has four components a Type of Utilization m Queue Distribution a Miscellaneous Performance Metrics a Top CPU Consumer HPB1787 B 04 00 HP
25. application definitions provides a rapid and simple means to select groups of programs or users to view in the Filter Display and Workload Screens GlancePlus looks for application definitions in the configuration file GLANCNFG when it is initiated If found GLANCNFG is used as the default application definition file If no application definitions are found there GlancePlus next looks for them in the PARM SCOPE SYS file This is the file used by HP Performance Collection Software which is part of both the HP LaserRX MPE and HP GlancePlus Pak products If no application definitions are found in either place then GlancePlus initiates with no default application definition file and the APPL Defn File field is blank To find out how to construct an application file see appendix D 5 28 Screen Metrics and Statistics Filter Display Screen The Filter Display screen provides both systemwide information and information that matches the programs users and applications you selected The Filter Display screen has five components m Program Banner Line a Global Bars a Filter Bars Summary Fields m Filter Processes The Program Banner Line and Global Bars are the same on all HP GlancePlus screens and contain the same information HPB1787 B 04 00 HP GlancePlus 13 02 25 LaserRX Parm File Filter a gOgOO OOOO OOOOO current avg high CPU M Ms sI ip BPs Be ccc cee eee 41 49 76 Disc MMS SI IB Be eee Ke
26. by the text running program on behalf of MPE iX If you wish to analyze that process in more detail you can use the gt command to take you to the Process screen for that process identification number PIN Note If you do not have System Manager SM capability you will see the letters CI instead of the command string text Screen Metrics and Statistics 5 43 Disc Detail Screen Press the D command to see activity for the system disc drives as shown on the Disc Detail screen This screen lists performance information for each drive currently active HPB1787 B 4 HP GlancePlus 16 32 49 Disc Detail Ma current avg high CPU Ms SI X IP Poo 82 7 63 95 Disc MSL Koc ccc cee eee ee ee te eee e eee teens i 14 18 29 Mem M MC ES SD DE 100 100 1007 Service Reads Writes Queue Max Queue 11 0 0459 1 9 1 4 2 27 44 0 0477 0 5 7 z3 8 0 0449 1 3 0 4 2 16 9 4 8380 1 7 0 6 2 22 6 8 0373 1 3 0 4 1 17 15 REE 1 9 0 9 1 27 Logical Attempts Reads R lt 48 5 Writes oo oo o o lt 11 9 Memory Manager gt Reads ooooooo lt 1 1 Writes oo oo o o lt 1 9 Top DISC Consumer is Pin 306 at Z I 0 s per sec LOGCHK PUB SECURE Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 20 Disc Detail Screen The detailed data area of the display is divided into three parts a Individual Disc Utilization a Summ
27. new queue and as such is attributed to the application that matches the process at the end of each 1 minute sample interval D 6 Application Definition Files OR Use the OR parameter to allow more than one application definition to apply to the same application Within a single application definition a process must match at least one of each category of parameters Parameters separated by the OR parameter are treated as independent definitions If a process matches the conditions for any definition it will belong to the application as in the following example APPLICATION CONSOLE USER OPERATOR SYS OR USER MANAGER SYS FILE STORE PUB SYS SYSDUMP PUB SYS This defines an application CONSOLE that consists of any programs run by the user OPERATOR SYS plus two programs STORE PUB SYS and SYSDUMP PUB SYS if they are executed by the user MANAGER SYS Sample Application Definition The following is an example of how an application can be defined APPLICATION Program Development INTERACTIVE EDITOR PUB SYS QEDIT VOODOO TDP USER TEST TESTGROUP LABTEST 0 Q APPLICATION Other Editors FILE EDITOR PUB SYS QEDIT VOODOO TDP APPLICATION Compilers FILE COBOL PUB SYS FORTRAN PUB SYS FIN PUB SYS FILE PASCAL PUB SYS SPL PUB SYS RPG PUB SYS APPLICATION HP Products FILE PUB SYS Application Definition Files D 7 The following is an example of how programs would be logged using the preceding application d
28. on commands m Chapter 5 to learn about the statistics appearing on each screen m Chapter 6 to find out about the Expert facility m Chapter 7 for information on customizing HP GlancePlus The appendixes contain additional information such as expanded definitions of wait reasons HP GlancePlus also provides online help information as discussed in chapter 3 This online help information is context sensitive and can be accessed using the H command Introduction 1 3 Installation System Requirements HP GlancePlus for MPE iX can be used on any HP 3000 Precision Architecture system using the MPE XL operating system version 2 1 or later or the MPE iX operating system Terminal and PC Requirements HP GlancePlus runs on the following terminals and personal computers m HP Vectra personal computer running HP compatible terminal emulation software HP 150 personal computer HP 700 9 family HP 2392 family HP 2622 family Installation 2 1 Installing HP GlancePlus Installing GlancePlus is a simple process It involves placing the appropriate files in a group with the necessary capabilities Group Capabilities HP GlancePlus should be placed in a group that allows all system users easy access to the program Normally this group is PUB SYS The group must have the following capabilities m Interactive access IA m Process handling PH a Extra data segments DS m Privileged mode PM
29. own the HP LaserRX MPE or HP GlancePlus Pak programs application definitions will probably be found in the PARM SCOPE SYS file See the HP Performance Collection Software manual for more details 7 4 Customizing HP GlancePlus Since you can set parameters in several ways it is important to understand what happens if you use more than one way The hierarchy for setting parameter values is usually Program Defaults GLANCNFG File Job Control Words which can be overridden by which can be overridden in most cases by which can be overridden by Value Command Changes The following table lists the changeable parameters their default values and the key words used to set them Detailed descriptions of these parameters follow the table Table 7 1 Changeable Default Parameters Interesting Process Thresholds Values Job Control GLANCNFG File Default Command Word Keyword Value MinCPU GLANCEMINCPU THRESHOLD MINCPU 1 Percent CPU GLANCECPU THRESHOLD CPU 10 Percent DISC GLANCEDISC THRESHOLD DISC 10 1Os Second RESP GLANCERESPONSE THRESHOLD RESP 5 Seconds TRANS GLANCETRANS THRESHOLD TRANS 10 Trans Min NEW GLANCENEW THRESHOLD NONEW YES show DIED GLANCEDIED THRESHOLD DIED YES show WAITIMPEDE GLANCEWAITIMP WAIT IMPEDE 50 Percent WAITCPU GLANCEWAITCPU WAIT CPU 50 Percent WAITDISC GLANCEWAITDISC WAIT DISC 50 Percent WAITMEM GLANCEWAITMEM WAIT MEMORY 50 Percent Customizing HP GlancePlus 7 5 Display Control Parameters
30. size you combine the tates for the two modes and use table 5 2 as a guideline for determining whether excessive switching is occurring If it is you should consider a Native Mode migration Table 5 2 Guidelines for Switches to CM or NM Machine Threshold Rate Model switches second 5 920 approx 75 5 925 approx 120 5 922 approx 150 5 932 approx 225 5 935 approx 270 5 950 approx 300 5 955 917 927 937 947 approx 440 5 948 949 approx 500 5 960 approx 630 5 958 approx 650 5 957 approx 720 5 980 100 967 approx 1000 S 977 approx 1170 5 980 200 approx 1500 5 980 300 approx 1830 Screen Metrics and Statistics 5 41 Interval CM Interval Process Completions Current Ready Queue Maximum Ready Queue Launch Rate CQ SAQ Interval File Open Rate This shows the percentage of time the CPU was in Compatibility Mode between switches This shows the rate of process completions for the current interval number of processes completed per second This provides information about the workload characteristics of the machine A high rate of process completions for instance may indicate a CPU bottleneck due to excessive or improper process handling A sustained rate greater than 0 50 may be considered excessive This shows the number of processes waiting for the CPU Scale 0 to 25 processes This shows the maximum number of processes waiting for the CPU since GlancePlus was started or since the MPE
31. time that the process has been running CPU TOT CPU TOT shows the total amount of processor time used during the life of the process DISC TOT DISC TOT shows the total number of physical disc transfers completed during the life of the process A number followed by the letter K represents thousands of transfers a number followed by the letter M represents millions of transfers Note Disc totals are not collected until either HP GlancePlus or another performance tool enables the MPE measurement facility to collect process data This means that the DISC TOT value will not include physical transfers that were completed by a process before the MPE measurement facility was enabled Screen Metrics and Statistics 5 15 Process State Bars The bars show the process s states during its lifetime and the last interval State E ED DG OP PO 0 lt 1007 Last E EP PO 0 lt 1007 Figure 5 8 Process Screen Process State Bars STATE Components of Response Time The STATE or Elapsed Time Process State bar shows the components of response time for the process over its lifetime running time The number at the right represents what percentage of the total time the process requested service it was not waiting on terminal reads Usually this number will be 100 for a batch job since batch jobs typically do not interact with a terminal For interactive processes however this number is typically well below 100 and is a good indicator of how m
32. 0 0520 indicates an average service time of 52 milliseconds for that drive for the current interval The number of physical reads per second for the disc during the current interval appears to the right of the Service field The column showing the number of physical writes per second for the disc during the current interval appears to the right of the Reads column The Queue column lists the number of processes in queue for this disc during the current time interval The Maximum Queue column shows the largest number of processes in queue for this disc since GlancePlus or the measurement interface was started Screen Metrics and Statistics 5 45 Summary Disc Utilization The Summary Disc Utilization section shows the rate of logical reads and writes per second and the global memory manager reads and writes per second These rates are for the current interval Logical Attempt Bar scale 500 I Os per second Memory Manager scale 100 I Os per second Top Disc Consumer The bottom section of the Disc Detail screen lists the highest disc consumer for the current interval identified by its process identification number PIN the program name and the physical I O rate for that PIN If no disc activity occurred during the interval the Top Disc Consumer line will not appear on the screen If you wish to analyze that process in more detail you can use the gt command to take you to the Process screen for that PIN Logical Attempts
33. 27 time CPU 5 32 current 3 3 elapsed 5 16 5 31 response 5 16 5 18 service 5 45 7 16 total CPU 5 31 total disc 5 31 utilization 5 45 Time Adjust command 3 10 Timeout on prompts 4 4 top consumer CPU 5 34 5 36 5 43 disc 5 34 5 36 5 46 trace display 5 19 5 24 transaction parameter 7 8 transactions terminal 5 17 U UDC files 2 2 processes 5 34 RUN command 3 1 underlined metrics 5 3 6 2 7 16 Update Display command 3 10 usage Index 7 CPU 3 3 disc 3 3 memory 3 3 USER parameter D 5 Users command 3 10 4 3 V variables filter screen 5 26 user defined 7 17 WwW wait on CPU parameter 7 9 wait on disc parameter 7 10 wait on impede parameter 7 9 wait on memory parameter 7 10 WAIT reasons 5 17 components 5 6 5 11 detailed C 1 Index 8 expanded definitions B 1 WAIT states components 5 16 Workload command 3 10 Workload screen 5 34 36 interesting applications 5 35 top CPU consumer 5 36 top disc consumer 5 36 Xx Xpert Analysis command 3 10 3 12 6 5 Xpert facility 6 2 invoking 6 3 Xpert Overview command 3 10 6 4 Z Zero Averages command 3 7 3 10 4 4
34. 3 DIALUP 48 RESMNT 0 4 5295 334 DUP 13 DIALUP 103 REVALINQ 0 0 SZ97 333 DUP 41 DIALUP 239 TTSSON CPU 5298 225 HARBISON EVAL 144x 0 0 SON 5298 225 HARBISON EVAL 261 0 0 TERM Figure 5 2 Global Screen Interesting Processes Some interesting processes will have highlighted fields to indicate an event that might require closer attention These appear in a different color or are underlined on your screen depending on the type of monitor see figure 5 2 for examples of highlighted processes You can use the Set Option Values V command to change the threshold values of events to cause a process to become interesting or have one of its fields highlighted You can also change the threshold values before entering GlancePlus by setting the appropriate job control words JCWs See chapter 7 for more information on customizing GlancePlus Screen Metrics and Statistics 5 3 Interesting Process Components Each interesting process occupies a single line on the Global screen The components of interesting processes and any applicable reasons for highlighting are as follows JSNO JSNO is the job number Jnnnn session number Snnnn or process identification number P nnn for the process DEV DEV is the logical device number for a job or session If the job or session is a system process DEV is displayed as SYS LOGON LOGON is the first 22 characters of the logon string jobsessionname user account PIN PIN is the process identific
35. 5 21 5 27 5 29 5 34 5 37 5 38 5 39 5 39 5 43 5 44 5 45 5 46 5 47 Contents 5 5 24 Memory Detail Screen Fault Types 2 2 2 45 48 5 25 Memory Detail Screen Miscellaneous Metrics 5 49 6 1 Xpert Overview Screen o 6 4 Contents 6 Tables 3 1 Commands to Access Screens ne 3 10 3 2 Commands to Access Program Activities 11 5 1 Wait Reasons a 5 6 5 2 Guidelines for Switches to CM or NM o BA 7 1 Changeable Default Parameters AN 7 5 7 2 GlancePlus Highlighting Characteristics rr fe 06 7 3 GlancePlus Performance Indicators 020 7 19 C 1 Detailed WAIT Reasons 2 ee C 1 Contents 7 Introduction HP GlancePlus for MPE iX is an online diagnostic tool that you can use to monitor system activity and troubleshoot system performance problems What HP GlancePlus Can Do For You HP GlancePlus is easy to install easy to run and easy to use You don t have to spend time on complicated set up procedures or training Anyone can use HP GlancePlus effectively from day one HP GlancePlus is powerful It gives you a complete picture highlighting processes that are starved for resources and those using resources HP GlancePlus gives you expert assistance It uses a comprehensive set of rules developed by performance specialists to alert you whenever a possible performance problem arises HP GlancePlus is flexible You can customize the software t
36. 6 Global 3 9 3 10 Help 3 9 3 10 3 13 6 2 I O by Phys Disc 3 10 list of available 3 9 List Users 3 10 4 3 Memory Detail 3 10 MPE Command 3 10 3 15 4 2 5 4 5 11 5 25 Next Logical Display 3 10 4 1 4 4 parameters 3 1 Previous Display 3 10 4 4 Print 3 10 3 14 Refresh Screen 3 10 4 2 RUN 3 1 Select 3 10 4 1 Select Job 3 10 4 1 Select Process 3 9 3 10 4 1 Select Session 3 10 Set Option Values 3 10 3 16 5 3 5 19 5 20 6 5 6 7 7 4 single letter 3 2 3 10 Time Adjust 3 10 Update Display 3 10 Users 3 10 4 3 Workload 3 10 Xpert Analysis 3 10 3 12 6 5 Xpert Overview 3 10 6 4 Zero Averages 3 7 3 10 4 4 Commands Menu command 3 10 compatibility mode 7 16 components CPU bar 3 4 disc bar 3 5 memory bar 3 6 configuration parameters 5 19 configuring Expert facility 7 15 continuous display 5 20 conventions 1v capitalization iv computer font iv Enter lv italics iv Return iv uppercase 1v CPU overhead 3 7 parameters 7 7 usage 3 3 CPU bar 3 4 7 2 components 3 4 CPU Detail command 3 9 3 10 5 37 CPU Detail screen 5 37 43 miscellaneous performance metrics 5 39 queue distribution 5 39 top CPU consumer 5 43 type of utilization 5 38 customizing GlancePlus filter command 5 26 D data filtering 6 1 default parameters 7 5 detail display 3 3 detailed WAIT reasons C 1 dialogue filter 5 26 died param
37. 7 3 7 2 0 0 TERM DSZ3A RECUUSER PRODSY 761 ITS2Z2 5 P C152 08 17 0 1 Z 5 1 TERM SDBZZA RECUUSER PRODS 18Z9xSFC1B15P C152 1 9 0 1 24 0 1 TERM The lower section of the Filter Display screen provides details about processes that match the filter criteria These fields contain the following information JSNO DEV LOGON PIN PROGRAM PRI CPU This field contains the job number Jnnn or session number Snnn or if the process is a system process the PIN or process identification number Pnnnn as appropriate This field contains the logon device number SYS indicates a system process This field contains the user logon string This field contains the process identification number System processes will have the same PIN listed here and under JSNO This field contains the program s name or if the program is the MPE Command Interpreter CI then the field will display a colon followed by the first characters of the last CI command issued This field will contain the execution queue and priority number or if terminated the word DIED This field indicates the percentage of time the CPU was used during the last interval 5 32 Screen Metrics and Statistics DISC This field indicates the rate of physical disc transfers number per second during the last interval TRN This field indicates the rate of terminal transactions number per minute during the last interval RESP This field indicates the ter
38. BMS wait state The process is waiting for the completion of nondisc I O such as from magnetic tape Disc I O DISC and terminal I O TERM have separate wait states and are not included in this category B 2 Wait Reasons Expanded Definitions MEM MISC MSG OTHR PAUS RIN The process is waiting for library code program code data or any combination of the three to be brought into memory Since it would be wasteful for a large process to fully occupy main memory the process is broken up into smaller pieces referred to as pages As these segments or pages of memory are needed they are brought into main memory from disc and the process must memory MEM wait It is normal to find a process in this state occasionally Finding a single process in memory wait state repeatedly however can be caused by poor locality the process s logic is forcing it to continually branch to other segments or pages Finding multiple processes in this state can indicate a lack of main memory As the demand for memory outstrips the supply the operating system must use a process s main memory for other higher priority processes The process is waiting for an unspecified wait state and is accumulated under this general category HP GlancePlus reports seldom used wait states in this category The process is waiting to receive a message To synchronize internal operations some operating system processes communicate by sending internal m
39. CUTE access Aoptions Access 6 File is opened for EXECUTE READ Aoptions Access 7 File is opened for MAPPED FILE access The way the file is shared by this process EXC SEM File is open for EXCLUSIVE access Aoptions Excl 1 File is open for READ SHARE SEMI EXCLUSIVE access Aoptions Excl 2 File is open for SHARE access Aoptions Excl 3 File is open with Dynamic Locking Aoptions Lock 1 Screen Metrics and Statistics 5 23 Opens The number of times the file has been opened TurboIMAGE data sets will have their record pointers and end of files expressed as TurbolMAGE record numbers instead of MPE file blocks of media records See the MPE Intrinsics Manual under the FOPEN intrinsic for a detailed description of Foptions and Aoptions parameters Display Family Press the Display Family function key to see the family tree for the process currently being viewed To display the tree without using the function keys press 2 The 2 subcommand is valid only during the Process screen display If available the parent process identification number PIN is shown followed by any and all children of the process For each PIN displayed the program being executed by the respective process is also presented Display Trace Pressing the Display Trace function key provides an HP Debug Procedure trace of the process being monitored To display the trace without using the function keys press 3 The
40. For example if you are monitoring a batch job that terminates GlancePlus returns to the Global screen automatically MPE Command Use this command to enter a single command to be executed by MPE This includes MPE UDC RUN and implied RUN commands The HP GlancePlus screen is suspended until the command completes and you press Return or Ena Hint You can issue subsequent commands after a single command is completed by simply entering a colon in response to the prompt Press RETURN to continue or Enter a GLANCE command Refresh Screen R Use this command to clear the current display and redraw it with updated information for the current interval It is useful for solving the problem of a corrupted display by refreshing the display completely All Processes A Use this command to produce a one time screen display that lists all processes or applications not just those that are interesting It can be used with the Global Job Session Workload and Filter screens Following the All Processes display the screen returns to its previous method of displaying only interesting processes 4 2 Commands Additional Information Change Queue Q Use this command if you have sufficient capability to change the queue of a specific process or of all the processes associated with a specific job or session A user with system manager SM capability can also change a process or all the processes belonging to a job
41. GLANCEXPERT 1 for continuous VALUES XPERT CONTINUE or XPERT DEMAND The EXPERT LEVEL parameter controls whether the Expert facility is working at a low level of detail no process or at high detail including processes and slightly higher overhead See chapter 6 on the Expert facility for information on how to set this parameter DEFAULT LOW GLANCNFG XLEVEL LOW or XLEVEL HIGH JCW SETJCW GLANCELEVEL 1 for low or SETJCW GLANCELEVEL 2 for high VALUES XLEVEL LOW or HIGH Customizing HP GlancePlus 7 13 Additional Parameters The EXPERT USERS parameter can control which users are allowed to use the Expert facility who can execute X commands or utilize continuous mode DEFAULT ANY Any user may use the Expert facility GLANCNFG XPERT SM OP AM ANY or any combination to allow users with SM OP AM capabilities or ANY user to use the Expert facility XPERT NONE Prevents all users from accessing the Expert facility JCW This capability is not accessible through Job Control Words VALUES This capability is not accessible through any GlancePlus commands The QZAP parameter is only allowed in the GLANCNFG file It can control which users are allowed to execute the QZAP command to change the priority of a process job or session DEFAULT SM Only system managers may erecute the QZAP command GLANCNFG QZAP SM Only system managers may execute the QZAP command QZAP OP System operators or users with
42. HP GlancePlus User s Manual for MPE iX Systems a HEWLETT C PACKARD HP Part No B1787 90001 Printed in U S A 0492 Edition 2 E0492 Notice Hewlett Packard makes no warranty of any kind with regard to this material including but not limited to the implied warranties of merchantability and fitness for a particular purpose Hewlett Packard shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing performance or use of this material Hewlett Packard assumes no responsibility for the use or reliability of its software on equipment that is not furnished by Hewlett Packard Copyright 1992 Hewlett Packard Company This document contains proprietary information which is protected by copyright All rights are reserved No part of this document may be photocopied reproduced or translated to another language without the prior written consent of Hewlett Packard Company The information contained in this document is subject to change without notice Microsoft MS DOS and Windows are U S registered trademarks of Microsoft Corporation Hewlett Packard Company SWT Software Technology Center 8010 Foothills Boulevard Roseville CA 95678 U S A Printing History Updates are new editions or complete revisions of the manual A software code may be printed before the date this indicates the version level of the software product at the time the
43. KS 327 1007 C 07 C 07 C 07 CM 327 1007 C 07 C 07 C 07 SWITCH 327 1007 C 07 C 07 C 07 RESPONSE 327 1007 C 07 C 07 C 07 Figure 6 1 Xpert Overview Screen 6 4 The Expert Facility High level analysis can be performed only if the Expert facility is enabled for high level use the V command XLEVEL HIGH After a global analysis in which a problem type was not normal the processes that executed during the last interval are examined If an action can be suggested that might improve the situation the action is listed as follows XPERT Status 75 CHANCE OF GLOBAL CPU BOTTLENECK Reason INTERACTIVE gt 90 00 96 4 Action QZAP pin 122 PASXL for MEL EELKEMA from C to D queue Actions will not be instituted automatically since you may or may not agree with the suggestions If the Expert facility is enabled for high level analysis you can request additional information when you request an expert diagnosis using the X command Following the initial display you can type X again to display the process data analysis This analysis often includes the process or processes that are probably causing the problems as well as those that are suffering from the problems For example Enter the X command XPERT Status 75 CHANCE OF GLOBAL CPU BOTTLENECK Reason INTERACTIVE gt 90 00 96 4 Action QZAP pin 122 PASXL for MEL EELKEMA from C to D queue The Expert Facility 6 5
44. Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 14 Workload Screen Interesting Applications The Workload screen displays only applications that use CPU during the interval to allow you to concentrate on applications currently affecting performance Interesting applications scroll down as new entries are added to the top of the listing thus keeping the latest information just below the Global bars You can scroll the terminal display to examine previous applications and intervals again You can also press the A key to see a one time display of all applications 5 34 Screen Metrics and Statistics Some interesting applications will have fields highlighted to indicate an event that might require closer attention On colored monitors these highlights are in color on monochrome monitors they are underlined Since applications represent groups of processes the thresholds for CPU usage disc transfers and transaction rate are set higher than they are for single processes These values are set by highlight specifications in the GlancePlus configuration file GLANCNFG Since response time is an average for all processes in the application the response time threshold value used for single processes is also used for applications See chapter 7 for more information on customizing GlancePlus Interesting Application Fields The application bars and fields appear below the Global bars and present information about how
45. Response Time This field shows the state this process is in according to the MPE iX Dispatcher Valid states are Executing Ready Short Wait Long Wait and Null Executing The process is using the CPU Ready The process is waiting its turn for the CPU Short Wait The process is waiting for disc I O such as a code page fault or file page fault Long Wait The process is waiting for other I O activities such as a terminal read or terminal write SIR RIN Parent Wait and other semaphore type waits fall under this category Null Appears only if a process is being created or is dying This field shows the rate of terminal transactions transactions per minute completed during the last interval This field shows the percentage of CPU time the process spent in Compatibility Mode during the last interval This field shows the rate of Native Mode switches completed switches per second from Native Mode to Compatibility Mode during the last interval Bar Scale 0 to 200 switches per second The number to the right of the bar shows the actual switch rate per second up to a maximum of 9 999 This field shows the rate of Compatibility Mode switches completed switches per second from Compatibility Mode to Native Mode during the last interval Bar Scale 0 to 200 switches per second The number to the right of the bar shows the actual switch rate per second up to a maximum of 9 999 This field shows the average response time
46. abnormally high for this size of system this field is highlighted in the display GlancePlus s highlighting provides an expert in your pocket that can tell you when your system is out of line The Expert Facility 6 1 Note Highlighting works on all supported terminals and terminal emulators but it works best on color terminals Color highlights can provide additional information red indicates more severe situations than blue or green Terminals without color capability use underlining or inverse video to highlight an item Highlighting is not available on the printer output Guidelines GlancePlus provides guidelines to help you understand and use the data presented Usually all you have to do is type H or press 17 to access online Help Help can give you a list of commands and explain your current display in general terms Help can also give you more detailed information After entering the general Help section you can ask for details about each metric on your screen The advice of many HP 3000 performance specialists has been gathered and used to provide guidelines for understanding performance data By asking for GUIDELINES you can access this knowledge base for information on what is considered normal for your system and how to tell when your system might be experiencing problems This information is always at your fingertips when you need it a Expert Facility GlancePlus s Expert Facility helps you examin
47. age Overlay Candidate Rate 5 4 sec System Library Page Faults B 1 sec Memory Mgr Clock Cycle Rate 7 Physical Memory Size 256 mb Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 23 Memory Detail Screen Screen Metrics and Statistics 5 47 Fault Types The Fault Types section presents several metrics for memory faults faults per second Scale 0 to 50 faults per second System Memory Page Faults By Ob ject Type Transient E NN lt 14 7 sec Permanent ooocoooooocoooooccnooocccnoooccaooonooooooooooooooooo nao lt 4 B sec NM Stack N or lt 4 6 sec CM Stack E Coocnnoco lt Z 4WVsec NM Code coccoccccococcccooocccooocccoo ro lt 0 sec CM Code coccoccccocccccoocccoo ooo lt 0 sec CM Data ray lt 1 8 sec File Object FB cc cc ee eee eee eee teen eee eee eens lt 1 6 sec NM Sys Lib cee ee cc ce ccc cc cee ee eee ee ee ee eee ee eee eee eee eee eeene lt 0 sec CM Sys Lib Loc ee cc ccc cc ce eee ee eee ee eee eee eee eee eee eens lt 0 1 sec Figure 5 24 Memory Detail Screen Fault Types The Fault Types are as follows Transient Unspecified transient data objects Permanent Unspecified permanent data objects NM Stack Native Mode process stacks CM Stack Compatibility Mode process stacks NM Code Native Mode code CM Code Compatibility Mode code program files CM Data Compatibility Mode data segment File Object File Objects NM System Native Mode syste
48. arization of expert analysis is available when the Expert facility is used in continuous mode When this is enabled global analysis is performed after each screen update but messages are not given unless there is a problem When one or more problems are detected a single line summary is added to the current display screen In generating this one line summary each symptom is counted as one or more types of problems CPU DISC MEMORY LOCKS Compatibility Mode Switch Rate or Response Time The probabilities for all symptoms of a given type are added to arrive at the state of the resource If the state of the resource is low enough the resource is considered normal if higher it is evaluated as MODERATE HIGH or VERY HIGH The continuous mode Expert facility display is a single line showing any problem areas and their states For example XPERT CPU IS VERY HIGH SWITCHES ARE HIGH In addition to generating one line displays the Expert facility also retains a summary of how often each type of problem occurred in each state By typing 0 for Overall Summary you can generate a summary report showing the totals since the Expert facility was put into continuous mode perl Summary Gut of he Last S27 seconds 5 4 Minutes OK MODERATE HIGH VERY HIGH CPU 306 94 10 3 C 07 11 37 DISC 327 1007 C 07 C 07 C 07 MEMORY 283 877 44 13 C 07 C 07 LOC
49. ary Disc Utilization a Top Disc Consumer 5 44 Screen Metrics and Statistics Individual Disc Utilization The individual disc utilization section of the screen displays one line for each disc drive Information is shown on the logical device number Ldev percentage of utilization for the current interval the average service time Service physical read and write rates current queue depth and maximum queue depth Ldev Utilization Service Reads Writes Queue Max Queue 1 UU eee eee lt 11 0 0459 1 9 1 4 2 27 Z Ure c cee eee e ee ee ee lt 4 8 0477 0 5 0 4 2 23 3 Yo ee eee ee eee lt 8 8 0449 1 3 0 4 2 16 4 UU eee ee eee lt 9 0 8388 1 7 0 6 2 22 C EREEREER EERTETEEET lt 6 0 B373 1 3 0 4 1 17 9 UU ee eee eee lt 15 1 9 0 9 1 27 Figure 5 21 Disc Detail Screen Individual Disc Utilization Utilization Service Reads Writes Queue Max Queue The utilization for each disc drive refers to the percentage of time during the current interval that the drive was in use servicing reads and writes It is computed by dividing the disc s service time for the interval by the interval s elapsed time Scale 0 to 100 utilization The average service time for each physical I O appears in the Service field Represented in seconds this metric is derived by dividing the disc service time by the total number of physical reads and writes for the current interval For example
50. ation definitions Since the default access for ANY user to PARM SCOPE SYS is execute only GlancePlus will be able to read it as long as it has execute access Note that read access is required for GlancePlus to read GLANCNFG Application Definition Files D 1 Application Definition Syntax The following information describes the syntax used to define an application and gives several examples of application definitions APPLICATION application name FILE filename INTERACTIVE filename BATCH filename USER job session name username acctname groupname QUEUE execution priority OR where The application name parameter identifies an application that groups multiple program files together and reports on their combined activities The application name is a string of up to 20 characters identifying that application One or more application definitions should follow the APPLICATION command An application definition consists of combinations of FILE INTERACTIVE BATCH USER or QUEUE parameters The OR parameter is used to separate multiple definitions for the same application All application definitions apply to the last APPLICATION command entered D 2 Application Definition Files The parameters for application definitions fall into three categories File name FILE INTERACTIVE BATCH User logon USER Execution priority QUEUE If a parameter from a category is used in an app
51. ation number for the process PROGRAM PROGRAM refers to the name of the program being run If the program is the MPE iX Command Interpreter CT then this field displays a colon followed by the first characters of the last CI command issued However depending on your assigned capabilities the following might apply a SM System Manager capability will show the last CI command for all users systemwide AM Account Manager capability will show the last CI command for all users within that account m A user with neither SM nor AM capabilities will be shown only his or her last CI command All other Command Interpreters will show CI as the program name The program name of a new process is highlighted and an asterisk is located between the PIN and PROGRAM columns A terminating process is represented by the word DIED highlighted in the priority PRI field PRI PRI is the execution queue where A AS B BS C CS D DS E ES followed by the current execution priority number of the process 1 255 If a process has just terminated this field will contain the word DIED 5 4 Screen Metrics and Statistics CPU DISC TRN RESP WAIT CPU is the percentage of the central processing unit used by this process during the last interval The CPU field is highlighted if the process exceeds the CPU threshold The default is 10 percent of the CPU DISC is the disc transfer rate transfers per second for
52. be able to understand that it is reasonable for a two processor system to have a Global CPU bar showing 100 usage and two processes in the interesting process section each using 100 of a single processor The user must understand the difference between Global CPU bar and individual process presentations Once this is understood the user need not consider the number of CPUs in use to easily understand how CPU intensive each process is and how much total processing power is in use A 2 Multiprocessor Systems Wait Reasons Expanded Definitions A process will be suspended if it cannot access a resource or it must wait for another event to be completed Although a process s wait reason can pass through many different states between refresh intervals HP GlancePlus reports the state at the instant the process is displayed The following WAIT reasons occur for MPE iX BREK IMP PAUS CONS 1 0 RIN CPU MEM SEM DBMS MISC SIR DISC MSG SON FATH OTHR TERM The specific WAIT REASONS can be defined as follows BREK The process is waiting to be resumed from a BREAK For instance if a user who is running a program presses the Break key and receives the colon prompt that process will show up in a BREK wait CONS The process is waiting for a reply from the console operator A typical example would be a STORE with an outstanding tape request at the console Wait Reasons Expanded Definitions B 1 CPU DBMS DISC FATH IMP
53. c Fill Operation Ctr1lBlck Control Block Wait HL I0 Wait for High Level I O Son Wait for Child FSTermI0 Wait for File System Terminal I O DataComm Data Comm Wait MemMgrPW Memory Manager Post Wait Console RIT Console Reply Wait SignlImr Signal Timer Wait Dspatchr Wait for Dispatcher Disp Work Other Other Wait Reason Detailed WAIT Reasons C 1 D Application Definition Files The GlancePlus Workload screen and appropriate Filter commands allow you to utilize user defined applications These are the same definitions as those used by the HP LaserRX MPE and GlancePlus Pak for MPE iX products You can use any one of the following as an application definition file m GlancePlus configuration file GLANCNFG PUB SYS m LaserRX or GlancePlus Pak Parm file PARM SCOPE SYS m Any other file that contains valid application definitions GlancePlus scans the GLANCNFG file automatically every time it is executed If application definitions are not found it attempts to open PARM SCOPE SYS If you want to use a file other than the ones GlancePlus opens select it by using the APPL keyword in the Filter or Value command If no applications are defined in the GLANCNFG file you may file equate PARM SCOPE SYS to an application definition file of your choice before starting GlancePlus Then it will begin with the desired application definitions Read access is usually required before a user is allowed to read a file containing applic
54. c c ccc creer rr rr rr i z9 44 101 Mem M MC ES SD DE 100 100 1007 CPU O lt 6 1 C 4 Fred Disc FFPXo oooooooooo ooo ooooooo lt 9 8 Elapsed CPU Tot Disc Tot 7 2 hr i 6 8 min 18 7 k JSNo Dev Logon Pin Program Pri CPUZ Disc Trn Resp Wait s11 754 FFBZ3A RECVUSER PRODS 204 ITSZ 5 P C152 8 47 0 0 29 0 0 TERM J83 10 RPSZ000J RPSBATCH PRO 1529 RPSZ Q P D202 39 9 3 7 2 0 0 TERM S584 8 DSZ3A RECUUSER PRODSY 761 ITSZ 5 P C152 8 17 Q 1 Z 5 1 TERM S5290 435 SDB2ZA RECUUSER PRODS 1829 SFC1 15P C152 1 97 0 1 24 0 1 TERM Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 5 13 Filter Screen Screen Metrics and Statistics 5 29 Filter Bars CPU rr lt 6 1 i Disc F FR eee ec ee ee ee eee lt 9 8 l l CPU Bar The CPU bar represents the usage of CPU resources for all processes filtered during the last interval The F code on this bar corresponds to the Filter Display screen and may indicate jobs sessions or system processes that match the filter criteria you selected The asterisk indicates the average usage of CPU resources by the filtered processes for the total time monitored but only since the last Z command if one was entered See chapter 4 for information about screen commands Disc Bar The disc bar shows the combined rate of disc transfers for all processes filtered during the last interval The F code indicates the job session and system process
55. ceptionally long to use them and are the processes most likely to reflect an existing or potential problem The procedure for setting process thresholds is described at the end of this chapter The Global screen is the starting point for most performance analysis activity Other GlancePlus screens are discussed individually in chapter 5 HP GlancePlus Function Keys and Commands To make it easy to select screens or functions GlancePlus displays a row of function key labels across the bottom of each screen These labels also called softkeys correspond to function keys through on your terminal keyboard Global Select Print MPE Time Refresh Help Exit Command Ad just Sereen Glance Figure 3 5 HP GlancePlus Function Keys You can also access all of GlancePlus s functionality using the single letter commands shown in tables 3 1 and 3 2 see pages 3 10 and 3 11 To avoid confusion function keys perform the same actions on all GlancePlus screens To maintain this consistency some screens or functions can be selected only by using their single letter commands The command abbreviations such as C for CPU screen and D for Disc screen are intuitive and easy to remember For a quick reminder simply enter H or press the Help softkey then press C at the next prompt to indicate you want a quick list of the available commands Additional information about commands is presented in chapter 4 Getting Started 3 9 Table 3 1 Commands to Acce
56. disc transfers that occurred on the system during the last interval The current average and highest number of transfers per second are shown numerically to the right of the bar The length of the bar indicates the total number of transfers per second occurring currently Scale 0 to 100 physical transfers per second The components of the Disc bar are as follows M S Memory management transfers swapping System process transfers data communication monitors and spoolers Interactive session process transfers Batch job process transfers Average usage of this resource during this execution of GlancePlus Getting Started 3 5 Memory Bar CPU ff AI IBP ee IN S2 66 89 Disc I A i 6 14 29g Mem amp ME ES SD DE 3 100 100 1007 This bar represents the amount of main Memory in current use as a percentage The current average and highest percentages of main Memory use are displayed to the right of the bar The length of the bar indicates the total amount of main Memory that is currently being used Note It is normal for main Memory to be full even when there is little activity on the system Components of the Memory bar show how much Memory is being used for different purposes The Memory bar components are as follows M Resident MPE operating system code and data are not swapable C Code both native and compatibility mode programs and libraries S Stacks both native and compatibility mode
57. e all the pertinent system performance indicators quickly and obtain a diagnosis It can even recommend specific actions that might improve your system s performance This facility is a true expert system utilizing advanced techniques that enable this program to make the same decisions as might be made by a human expert in HP 3000 computer system performance 6 2 The Expert Facility Invoking The Expert Facility You can use the Expert facility in several ways m It is always active and can guide you when fields are highlighted on a specific GlancePlus display m You can type X from any screen to request the Expert facility to diagnose your system This diagnosis considers all available information not just the information displayed on the screen mw You can enable the Expert facility to operate continuously In continuous mode the Expert facility examines your system each time the display is updated If it notices any performance problems it will add a single line summary to the screen indicating the area of the problem and its severity The Expert facility will also notify you of any problems regardless of the screen you are viewing the only exceptions are the Help Option Values and MPE Command screens What Does The Expert Facility Do The data displayed on each GlancePlus screen is examined by the Expert facility and any indicators that exceed the normal range for the size of system are highlighted Since the highlightin
58. e current record pointer for the file This usually points to the last record accessed by the process End of File The current end of file value disc files only Domain The designation of the file s condition It may be NEW File is newly created Foptions Domain 0 PERM File is permanent Foptions Domain 1 TERM File is temporary Foptions Domain 2 OLD File is old permanent or temporary Foptions Domain 3 5 22 Screen Metrics and Statistics Access Share If the file is a system defined file the Domain lists one of the following designations stdlist newpass oldpass stdin stdinx null 2777777 The way a follows IN OUT KEEP APEN 1 0 UPDT XEQ LOAD M System defined file Foptions designator 1 System defined file Foptions designator 2 System defined file Foptions designator 3 System defined file Foptions designator 4 System defined file Foptions designator 5 System defined file Foptions designator 6 System defined file Foptions designator 7 file is opened by the process the type of access as File is opened for READ access Aoptions Access 0 File is opened for WRITE access Aoptions Access 1 File is opened for WRITE SAVE access Aoptions Access 2 File is opened for APPEND access Aoptions Access 3 File is opened for READ WRITE access Aoptions Access 4 File is opened for UPDATE access Aoptions Access 5 File is opened for EXE
59. e each key word s unique in the Filter Selection menu the program can recognize a key word by the first letter of its name a Multiple filter commands can be entered by separating the individual commands with semicolons Commands that have only two states such as YES no can be entered without a value to change to the opposite state For example if use THRESHOLD values YES no is YES it can be changed to NO by entering the THRESHOLD command m To remove a program name or user logon selection enter the key word without a value 5 26 Screen Metrics and Statistics An example of a filter selection menu is shown below Filter Selection Criteria keywords are UPPERCASE use THRESHOLD values YES no MATCH criteria all ANY PROGRAM name USER logon APPL Defn File PARM SCOPE SYS 16 Applications defined 1 OTHER Z Networking 3 Corp Utility 4 EDC 5 MM 3000 6 DBS 7 Batchnet 8 Obsolete Sess 9 Netbase 10 Procurement 11 Fred 12 Prodsys 13 Partsid 14 Queries amp Reporti 15 Promis 16 System Support CHANGE lt keyWword newvalue gt Figure 5 12 Filter Screen Selection Dialogue You can choose from the following variables use THRESHOLD YES no Select Yes and GlancePlus will use the threshold values values to display only interesting processes Select No to display all processes interesting or not that match the filter criteria MATCH criteria all ANY Select all if a process must matc
60. e in 15 to 30 seconds After creating and activating the SPT program GlancePlus resumes normal operation and its screens can be accessed Note HP GlancePlus limits SPT data collection to one active process at a time Screen Metrics and Statistics 5 25 Filter Selection Menu Variables HP GlancePlus lets you define any collection of programs applications or users as a group and view them on the Filter Display screen It combines the ability to subset the Global screen s interesting processes to a group you wish to see with the Job Session screen s feature of summarizing their activity You can press the Filter command key E to see the current filter values You will be prompted to change them or you can accept the current values by pressing RETURN Also you can select whether to display information by program user or predefined application and whether it must match one or all of those selections to become a part of the filter group You can select whether it should display all processes matching your specifications or only those that exceed threshold values MPE wildcard attributes are accepted for the Program name and User logon To change a current filter s value type the key word for the filter shown in uppercase on the screen then type the new value for that filter Hints a HP GlancePlus can recognize a key word by the minimum number of characters needed to identify it uniquely In this case sinc
61. e number of processors Screens that represent groups of processes the Job Session Filter and Workload screens present their data as follows CPU bars and their associated percentages are displayed relative to all processors CPU seconds are displayed relative to a single processor For example a Job screen showing a job with one process constantly using the CPU would display the JOB CPU bar at 50 just like the Global bar The CPU total seconds would increment at the same rate as clock time The process in this display would still show CPU usage at 100 since it is displayed relative to a single processor These choices of design for presenting CPU usage represent tradeoffs It is useful to present process CPU usage relative to a single processor since Multiprocessor Systems A 1 other MPE sources such as the system log files the REPORT command the SHOWME command and log off statistics also present CPU usage relative to a single processor It is advantageous to represent CPU seconds on the Job Session and Filter screens relative to a single processor so they can be compared properly to other MPE sources In addition it is useful to scale the CPU bars between 0 and 100 of the total processing power of the system or relative to all processors By doing this 100 means that the total CPU power of the system is utilized completely regardless of the number of processors utilized The tradeoff for these advantages is the need to
62. e of display is used for the GLOBAL CPU bar at the top of each screen INTER BATCH divides the CPU bar by the type of processes using it SYSTEM INTERACTIVE and BATCH while QUEUE divides the CPU bar based on the priority of the processes A B C D E DEFAULT INTER BATCH GLANCNFG DISPLAY CPU INTERBATCH or DISPLAY CPU QUEUE JCW SETJCW GLANCEDISPCPU 0 for inter batch or SETJCW GLANCEDOSPCPU 1 for queue VALUES DISPCPU INTER BATCH or DISPCPU QUEUE The DISPLAY MEMORY parameter determines what type of display is used for the GLOBAL MEMORY bar at the top of each screen CONTENTS displays how much of main memory is occupied by different types of objects FAULTS shows the rate at which different objects are being requested to be brought into main memory faulted into memory DEFAULT CONTENTS GLANCNFG DISPLAY MEMORY CONTENTS or DISPLAY MEMORY FAULTS JCW SETJCW GLANCEDISPMEM 0 for contents or SETJCW GLANCEDISPMEM 1 for faults VALUES DISPMEMORY CONTENTS or DISPMEMORY FAULTS 7 12 Customizing HP GlancePlus The Expert Parameters The Expert parameters control the operation of the Expert Facility The EXPERT MODE parameter controls whether the Expert facility examines the system automatically every time the display is updated continuous mode or is available only after typing the X command on demand DEFAULT DEMAND GLANCNFG XPERT DEMAND or XPERT CONTINUOUS JCW SETJCW GLANCEXPERT 0 for demand or SETJCW
63. e of the job or session see note A number followed by the letter K represents thousands of transfers a number followed by the letter M represents millions of transfers Note Disc totals are not collected until either HP GlancePlus or another performance tool enables the MPE measurement facility to collect process data This means that the DISC TOT value will not include physical transfers that were completed by a process before the MPE measurement facility was enabled Upon entry to the Job Session screen CPU TOT and DISC TOT values represent CPU and disc totals for the processes associated with the job or session that are still alive To prevent these values from decrementing as processes terminate HP GlancePlus maintains counters for the job or session as long as it is monitored in the Job Session screen These counters are lost as soon as you enter another screen or select a different job or session Screen Metrics and Statistics 5 9 Interesting Processes Interesting processes related to the job or session you select appear below the Job Session Bars and Summary Fields In addition you can use the SYS option of the Select command to display all system processes not related to a job or session Pin Program Pri CPUZ Disc Trn Resp Wait Elapsed CPU Tot Disc Tot P Pin 144 REVALINQ C152 1 2 5 9 1 6 TERM 16 1 s 1 1 s 15 0 303 Figure 5 4 Job Session Screen Interesting Processes Interesting Job or Processes belo
64. e of total Time CPU was idle and disc IO did not occur percentage of total Total physical disc IO rate IOs second Physical disc IO rate by system processes IOs second Physical disc IO rate by interactive sessions processes IOs second Physical disc IO rate by batch job processes IOs second Physical disc IO rate to manage main memory IOs second Logical disc IO rate IOs second Average usage of all disc drives percentage Rate of terminal transactions completed transactions minute Average time to first terminal response seconds Average time to terminal prompt seconds Customizing HP GlancePlus 7 19 Table 7 3 GlancePlus Performance Indicators continued Indicator Description PEAK LOGL IO Logical disc IO rate for busiest disc drive IOs second PEAK PHYS 10 Physical disc IO rate for busiest disc drive IOs second PEAK MEM 10 Physical disc IO rate for memory manager on busiest disc drive IOs second PEAK UTIL Usage of busiest disc drive on system percentage of total CPU QUEUE Average number of processes using or waiting for CPU DISC QUEUE Average number of processes waiting on disc I O MEMORY QUEUE Average number of processes waiting on main memory IMPEDE QUEUE Average number of processes waiting on locks impedes etc CM CPU Average time spent in compatibility mode percentage NM CPU Average time spent in native mode percentage CM SWITCHES Rate of switches from native mode t
65. eatures added to the operating system to support POSIX for example hierarchical directories You may encounter references to MPE V the operating system for HP 3000s that are not based on the PA RISC architecture MPE V software can be run on PA RISC HP 3000s series 900 in compatibility mode Contents 1 Introduction What HP GlancePlus Can Do For You 1 1 How You Can Use HP GlancePlus 2 1 2 Will You Need Other Performance Tools 1 2 How to Use This Manual ee o 1 3 2 Installation System Requirements o 2 1 Terminal and PC Requirements 2242 2 1 Installing HP GlancePlus ee 2 2 Group Capabilities 2 o o 2 2 Group Access Lo a 2 2 Installing the Program Co 2 2 3 Getting Started Overview Lo 3 1 Running HP GlancePlus for MPE iX Lo 3 1 Exiting HP GlancePlus re 3 2 More Than a Look The HP GlancePlus Global Screen o 3 2 Reading the Screen Display 24242 3 3 Banner Line 3 3 Global Bars 2 o 3 3 CPU Bar 3 4 Disc Bar 3 5 Memory Bar 3 6 Percentage Columns o 3 7 Detail Display a 3 8 HP GlancePlus Function Keys and Commands Loe 3 9 Online Expert Assistance for Understanding Performance Data 3 12 Help Facility H 2 ee 3 13 Contents 1 Printing a Screen 2 2 7 ee ee LA Setting Process Thresholds 2
66. efinition Syntax a a a D 2 FILE So D 4 INTERACTIVE o D 5 BATCH ee D 5 USER 2 ee a D 5 OR Lo D 7 Sample Application Definition Lo D 7 Index Contents 4 Figures 3 1 3 2 3 3 3 4 3 5 3 6 3 7 5 1 5 2 5 3 5 4 5 5 5 6 5 7 5 8 5 9 5 10 5 11 5 12 5 13 5 14 5 15 5 16 5 17 5 18 5 19 5 20 5 21 5 22 5 23 The Global Screen The Banner Line Global Bars An Example of a Detail Display HP GlancePlus Function Keys An Example of the Expert Facility Setting Process Thresholds Global Screen Global Screen Interesting Processes Job Session Screen Job Session Screen Interesting Processes Process Screen Process Screen Process Bars Process Screen Process Summary Fields Process Screen Process State Bars Process Screen Miscellaneous Process Information Process Screen Screen Function Keys File Display Screen Detailed Display Filter Screen Selection Dialogue Filter Screen Workload Screen CPU Detail Screen o CPU Detail Screen Type of Utilization o CPU Detail Screen Queue Distribution CPU Detail Screen Miscellaneous Metrics CPU Detail Screen Top CPU Consumer Disc Detail Screen Disc Detail Screen Individual Disc Utilization Disc Detail Screen Summary and Top Disc Consumer Memory Detail Screen 3 2 3 3 3 3 3 8 3 9 3 12 3 16 5 2 5 3 5 7 5 10 5 13 5 14 5 15 5 16 5 17 5 19
67. efinition file Program User Logon Application EDITOR PUB SYS GERRY TEST Program Development EDITOR PUB SYS SAM SNEED Other Editors HPSLATE PUB SYS SAM SNEED HP Products PASCAL PUB SYS GERRY TEST Compilers EDITOR UTIL SYS GERRY TEST OTHER If the HP Products application in the previous example was defined first the first four examples would belong to it D 8 Application Definition Files Index A aaa key 3 10 4 1 ern 3 10 3 15 4 2 5 4 5 11 5 25 aaa lt key 3 10 4 4 aaa gt key 3 10 4 1 4 4 aaa key 3 13 aal A key 3 10 4 2 aa C key 3 9 3 10 5 37 aa D key 3 9 3 10 aa E key 3 2 3 10 aa F key 3 10 5 26 aal G key 3 10 a key 3 9 3 10 3 13 6 2 aa 1 key 3 10 aa J key 3 10 41 aa L key 3 10 aal M key 3 10 aa O key 3 10 6 4 aa P key 3 10 3 14 4 1 aa Q key 3 10 4 3 aa R key 3 10 4 2 aa S key 3 10 41 aa T key 3 10 Ho key 3 10 4 3 aa v key 3 10 3 16 5 3 5 19 5 20 6 5 6 7 7 4 aa w key 3 10 aal X key 3 10 3 12 6 3 6 5 aa Z key 3 7 3 10 4 4 a ctri Q key 3 10 GE key 3 10 a Ctrl HY key 3 10 additional parameters 7 6 7 14 All Processes command 3 10 4 2 Aoptions 5 21 application definition files 5 27 applications interesting 5 34 bars 5 35 fields 5 35 a Return key 3 10 a Space Bar key 3 10 AUTOINST process 2 2 available commands list 3 9
68. emory Detail screen 5 47 50 fault types 5 48 miscellaneous 5 49 memory usage 3 3 metrics color emphasized 5 3 6 2 7 16 highlighted 6 1 6 2 7 16 underlined 6 2 7 16 minimum CPU parameters 7 7 mode compatibility 5 18 7 3 7 16 native 5 18 7 3 7 16 MPE Command 3 10 3 15 4 2 5 4 5 11 5 25 multiprocessor systems A 1 N native mode 7 16 new parameter 7 8 Next Logical Display command 3 10 4 1 44 NOCCTL 3 15 O on demand display 5 19 online help 1 3 operating system software 2 1 OR parameter D 7 overhead 3 7 P page faults 7 16 rate 5 50 parameters additional 7 6 7 14 clear screen 7 12 CPU 7 7 default changeable 7 5 died 7 9 disc 7 7 display control 7 6 7 11 display CPU 7 12 display memory 7 12 Expert facility 7 6 7 13 expert mode 7 13 expert users 7 14 file detail display 7 11 hierarchy 7 5 minimum CPU 7 7 new 7 8 process keys 7 11 QZAP 7 14 QZAPSYS 7 14 refresh time 7 15 response time 7 8 RUN command 7 1 setting 7 5 transaction 7 8 wait on CPU 7 9 wait on disc 7 10 wait on impede 7 9 wait on memory 7 10 PC requirements 2 1 performance indicators 6 1 7 17 other tools 1 2 recommendations 7 17 Previous Display command 3 10 4 4 Print command 3 10 3 14 Print facility 3 14 process bars 5 14 components 5 14 processes count 5 8 5 31 5 35 details 3 8 display 3 3 Index 5 i
69. es that match the filter criteria you selected The asterisk on this bar indicates the average rate of disc transfer for the filtered processes over the total time monitored but only since the last Z command if one was entered See chapter 4 for information about screen commands 5 30 Screen Metrics and Statistics Summary Fields i lt 4 Fred Elapsed CPU Tot 7 2 hr i 6 B min The summary fields for the user defined filter appear at the right of the CPU and disc bars These fields contain the following information Count This number in parentheses is the total number of processes displayed There are four processes in the example above Filter Value This value identifies the filter It may be a program name user logon application name or application number list The example above identifies an application called Fred Elapsed This field shows the total time that the filter processes are known to have been running CPU TOT This field shows the total CPU time used by filtered processes during the time they were running DISC TOT This field shows the total number of disc transfers completed by all filtered processes during the time they were running Screen Metrics and Statistics 5 31 Filter Processes JSNo Dev S11 754 J83 10 s580 808 S5290 435 Logon Pin Program Pri CPUZ Disc Trn Resp Wait FFBZ3A RECUUSER PRODS 284 ITS2Z2 5 P C152 0 47 0 0 29 0 0 TERM RPS2 00J RPSBATCH PRO 1529 RPS2Z P D202 9 9
70. esh Help Exit Command Ad just Sereen Glance Figure 3 1 The Global Screen You can access other GlancePlus screens using the softkeys dedicated function keys or single letter commands described later in this chapter 3 2 Getting Started Reading the Screen Display Although all GlancePlus screens appear similar they carry different information about system resources Each screen is divided into four sections Banner Line Global Bars Detail Display and Function Keys The Banner Line and Global Bars present the same information on every screen whereas the Detail Display section differs from screen to screen Banner Line The Banner Line at the top of the screen contains information on the version of the product product name current time hour minute second and screen identifier figure 3 2 The screen identifier may contain information pertinent to the particular screen such as the particular job number in the Job screen Figure 3 2 The Banner Line Global Bars Three Global Bars appear directly below the Banner Line figure 3 3 These bars and the percentages listed to the right of them show the system s use of three resources CPU Disc and Memory The bars are present on every screen to help you keep track of the overall global picture The letters superimposed on each bar specify how that portion of the resource is being used CPU ff AI IBP ee IN S2 66 89 Disc I A i 6 14 29g Mem a
71. ess thresholds cause too many or too few processes to display on the Global screen you might want to adjust the thresholds to match the activity on your system You can change GlancePlus parameters in several ways Some ways are local to a particular execution of GlancePlus such as using the V Option Values command to change interesting process thresholds Other ways carry forward from one execution to another within the same job or session You can set job control words JCWs to have GlancePlus remember settings within the same job or session You can also set a JCW using the V command to allow you to run GlancePlus again from the same job or session using those values You can also set JCWs external to GlancePlus for example GLANCE user defined commands UDCs can start GlancePlus with specific values each time the UDC is used You can set the default GlancePlus parameter values for the entire system by changing the GLANCNFG configuration file If a file named GLANCNFG already exists in the same group and account as the GLANCEXL program it will be examined for default parameters You will be allowed to set some parameters only via the GLANCNFG file parameters such as controlling which users may execute the QZAP command The GLANCNFG file contains additional information such as the global rules for the Expert facility and sometimes the definitions for user defined applications used by the Filter and Workload screens If you also
72. essages to one another MPE message file waits RECEIVEMAIL intrinsic waits and interprocess communication IPC waits are examples of user generated message wait state The process is waiting for an unspecified wait state called OTHR which excludes all of the other wait states OTHR appears as an O in the wait reasons and the wait state bar in the Process display The application program calls the PAUSE intrinsic which delays the program for a user specified length of time The program is currently waiting for this time to elapse before continuing Specific resources can be shared by a set of processes MPE provides users with a RIN resource identification number facility to prevent two processes from using a shared resource at the same time When one process locks a specific RIN a second process wanting to lock the same RIN will wait A RIN wait applies to both global and local RINs RINs are also used to lock MPE iX files Wait Reasons Expanded Definitions B 3 SEM SIR SON TERM The process is waiting for a software semaphore or flag other than those specified by SIR and RIN wait categories To maintain consistency certain operating system functions are performed serially by a single process To enforce this access pattern a software resource called a SIR is obtained before the critical function is performed SIRs are internally used MPE resources and are not available to users directly The father process is waiti
73. etail Display Some screens show a summary of processes such as the Global Job Session and Filter screens Interesting applications are scrolled down as new entries are added to the top of the listing to keep the latest information just below the Global bars The process information for the most current interval is displayed at the top of the detail area immediately below the global bars You can use terminal memory to scroll back to process data from previous intervals Other screens show statistics for an individual resource or individual process These screens refresh the detail display at every interval Examples of this type of screen include the CPU Disc Memory and Process screens The Filter Select and Set Option Values screens help you set parameters Their detail display is a table of value settings followed by a prompt you can use to change a value Every time you change a value a new adjusted table of values is displayed below the previous one until you terminate the dialogue by pressing Return 3 8 Getting Started On many systems it could take several screens full of data to display all currently running processes Evaluating that much information could be confusing To avoid this GlancePlus lets you limit the processes displayed to those that are interesting to you processes that surpass predetermined threshold usage values Interesting processes either use a major portion of the system s resources or must wait ex
74. eter 7 9 disc parameter 7 7 usage 3 3 utilization 5 45 disc bar components 3 5 Disc Detail command 3 9 3 10 Disc Detail screen 5 44 46 individual disc utilization 5 45 summary disc utilization 5 46 top disc consumer 5 46 dispatcher queue 7 16 display continuous 5 20 on demand 5 19 display control parameters 7 11 display control parametersl 7 6 display CPU parameter 7 12 display family 5 19 5 24 display files 5 19 5 20 5 21 display memory parameter 7 12 display trace 5 19 5 24 E equipment hardware 2 1 software 2 1 Exit Glance command 3 2 3 10 exit program 3 2 3 10 Expert Analysis command 3 10 6 5 Expert command 6 3 Expert facility 6 2 configuring 7 15 invoking 6 3 parameters 7 6 7 13 expert mode parameter 7 13 Expert Overview command 3 10 6 4 expert users parameter 7 14 F family display 5 19 5 24 fault types 5 48 file detail display parameter 7 11 file display 5 19 5 20 5 21 file opens 5 21 7 16 rate 5 42 FILE parameter D 4 file parameters 5 19 filter dialogue 5 26 processes 5 32 value 5 31 Filter command 3 10 5 26 Filter Display screen 5 29 33 Index 3 filter bars 5 30 filter processes 5 32 Summary fields 5 31 filtering 6 1 Filter screen filter dialogue 5 26 variables 5 26 Filter Select screen 3 8 Foptions 5 21 function keys 3 3 3 9 command list 3 9 global 3 9 Process screen 5 19 select process 3 9 functions process spec
75. for the process during the last interval If no terminal transactions were completed the response time value is zero Scale 0 00 to 60 00 seconds 5 18 Screen Metrics and Statistics Process Screen Function Keys When you display the Process screen the softkeys at the bottom of the screen change to functions related to the process being viewed By pressing these softkeys you can display the information described below If you choose not to use softkeys you can access GlancePlus functions using the subcommands described with the specific functions Display Display Display SPT Other Refresh Help Exit Files Family Trace Collect Keys Screen Glance Figure 5 10 Process Screen Screen Function Keys Setting Option Values V The process specific functions can be customized using the V command to set the following configuration parameters m File Simple or Detailed This option controls the information displayed for open files See the Display Files option below for more information m Prockeys Demand or Continuous This option controls how information on the three process specific functions Display Files Display Family and Display Trace are displayed o Demand When Demand is selected the functions are treated as requests to display the appropriate information once For example if Prockeys Demand and the Display Files function is selected from the Process screen the open file information is displayed once but
76. g feature adds negligible overhead it is permanently enabled A global system analysis is performed based on data obtained from a single sample This can be a response to an on demand request you pressed the X key or might occur automatically following each screen update if the Expert facility is in continuous mode During global analysis all pertinent systemwide performance indicators are passed through a set of rules These rules were developed by top performance specialists working on the HP 3000 The rules were further refined through use on a variety of systems of all sizes and configurations The response to these rules establishes the degree of probability that any particular performance situation called a symptom could be true If the analysis is performed on demand any symptom that has a high enough probability of being true is listed along with the reasons rules why it is probably the case as in the following example The Expert Facility 6 3 XPERT Status 75 CHANCE OF GLOBAL CPU BOTTLENECK Reason INTERACTIVE gt 90 00 96 4 This says that most experts would agree that the system is experiencing a problem when interactive users consume more than 90 of the CPU Currently interactive use is 96 4 Since the probability is only 75 not 100 some additional situations are not true In this case the number of processes currently starved for the CPU might not be high enough to declare a real emergency Summ
77. g the Screen Display in chapter 3 The number at the right of the bar is the percentage of time the CPU is busy not counting idle time or paused for disc time Scale 0 to 100 of total time busy An asterisk represents the average usage of this resource over time since the program started or since the last Z zero command The Disc bar shows the disc transfer rate transfers per second by the process during the current interval The codes on this bar correspond to the codes used in the Global Disc bar see Disc Bars under Reading the Screen Display in chapter 3 Scale 0 to 100 disc transfers per second An asterisk represents the average usage of this resource over time since the program started or since the last Z zero command 5 14 Screen Metrics and Statistics Process Summary Fields The Process Summary Fields at the right of the screen provide additional information about the process i TAS3090P TASPRG MMB9IREL Elapsed CPU Tot Disc Tot 115 5 min 7 1 min 3 2 k Figure 5 7 Process Screen Process Summary Fields The Process Summary Fields are as follows PROGRAM TYPE The type of program is displayed as NMPRG native mode CMPRG compatibility mode or PROG code in LISTF or OCTed one or more SOMs have been run through the Object Code Translator PROGRAM The fully qualified program name is displayed as FILE GROUP ACCOUNT ELAPSED ELAPSED shows the total clock
78. ghlight processes that died SETJCW GLANCEDIED 1 Specifies to not highlight processes that died SETJCW GLANCEDIED 0 Specifies to highlight processes that died DIED YES or DIED N0 YES means to highlight processes that died The WAIT ON IMPEDE parameter determines when a process should be displayed and its WAIT field highlighted due to excessive waiting on software locks latches and other synchronization events DEFAULT GLANCNFG JCW VALUES 50 percent WAIT IMPEDE percent SETJCW GLANCEWAITIMP percent WAITIMPEDE percent The WAIT ON CPU parameter determines when a process should be displayed and its WAIT field highlighted due to excessive waiting on access to the CPU DEFAULT GLANCNFG JCW VALUES 50 percent WAIT CPU percent SETJCW GLANCEWAITCPU percent WAITCPU percent Customizing HP GlancePlus 7 9 The WAIT ON DISC parameter determines when a process should be displayed and its WAIT field highlighted due to excessive waiting on data transfers to or from disc drives DEFAULT 50 percent GLANCNFG WAIT DISC percent JCW SETJCW GLANCEWAITDISC percent VALUES WAITDISC percent The WAIT ON MEMORY parameter determines when a process should be displayed and its WAIT field highlighted due to excessive waiting on main memory resources DEFAULT 50 percent GLANCNFG WAIT MEMORY percent JCW SETJCW GLANCEMEM percent VALUES WAITMEM percent 7 10 Customizing HP GlancePlus Display Control Parame
79. group of processes you are monitoring If you are displaying the Disc Detail screen you will go to the Process Screen of the highest Disc Consumer It provides you with a guided analysis capability that lets you quickly pinpoint and analyze the top resource consumer A second method uses the J S or P command character to select the job session or process to monitor For example if you enter an S GlancePlus will prompt you for a session number When you enter a valid session number the Session screen will be displayed with information about the requested session These three commands J S and P are mnemonic and require a single keystroke A third method uses the Select softkey or the command character Once the appropriate identifier and its number are entered the screen changes to show the details of a single job session or process Batch jobs and sessions are identified by the letter J or S followed by their job or session number Commands Additional Information 4 1 Jnnnn or Snnnn Processes are identified by the letter P followed by a process identification number or PIN Pnnn This method has the advantage of being available through a softkey It also gives you the capability to select operating system processes those not attached to a job or session by entering SYS when prompted for a job session or process number If no process matching the selection criteria can be found the user is returned to the Global screen
80. h all of the criteria or ANY if it must only match any one selection criterion PROGRAM name Select any valid MPE program name which may include wildcards For example GLANCEXL will list all current users of the HP GlancePlus program USER logon Select any valid MPE logon value which may include wildcards For example XYZ will list all users currently logged on to the XYZ account Screen Metrics and Statistics 5 27 APPL Defn File If an application definition file appears on the display with application index numbers and names below it see figure 5 12 you may immediately select applications to be displayed Select one or more applications by entering their index numbers separated by spaces or commas You may also use the or to add or subtract applications from the selected list When you select an application both its number and name are highlighted For example entering 2 5 in response to the CHANGE prompt in figure 5 12 would result in selecting and highlighting the Networking and MM 3000 applications Entering 6 12 in response to the next prompt would add DBS and Prodsys to the selected applications If no applications appear on the display or you wish to specify a different application definition file enter APPL filename in response to the CHANGE prompt If the file contains valid application definitions then the display will show the application index numbers and names from that file Creating a file with
81. he normal CS queue for sessions and DS and ES queues for batch 7 2 Customizing HP GlancePlus The CPU bar has the following components DISPCPU QUEUE Code Meaning M MPE processing memory management interrupts ICS etc A Processes running in the A subqueue system priority B Processes running in the B subqueue high priority C Processes running in the C subqueue normally interactive D Processes running in the D subqueue normally batch priority E Processes running in the E subqueue normally background P Paused for disc no CPU used but disc I O is occurring Blank CPU was idle and no disc I O was occurring Global Memory Bar Configuring the Global Memory bar to show page fault rates rather than contents might be useful on a system where memory is a bottleneck and page faults are used as key indicators to alert you of a problem The Memory bar has the following components DISPMEMORY FAULTS Page fault rate for different object types Code Meaning T Transient objects program heaps etc Permanent objects not normally faulted Data Stacks NM CM Code NM CM Data objects Files Library code NM CM nt oa aunn Y Customizing HP GlancePlus 7 3 Customizing Configuration HP GlancePlus for MPE iX comes ready to use Although it is configured with a selection of carefully tested defaults you might want to change those settings in certain situations For example if the interesting proc
82. her protect the file from unauthorized modification Configuring the Expert Facility Many parts of the GlancePlus Expert facility can be customized by editing the GLANCNFG file Be careful when making changes because they will affect the Expert facility s recommendations when you use it Two main areas of the Expert facility can be configured The highlighting of key metrics on GlancePlus screens and the performance recommendations made by the X command Customizing HP GlancePlus 7 15 Highlighting Metrics You can change the points where the Expert facility highlights metrics on GlancePlus screens by adding one or more lines to the GLANCNFG file These added lines should have the following syntax HIGHLIGHT hitem newvalue where hitem is one of the items shown in the following table and newvalue is the desired value for that item The hitem field must be entered exactly as shown uppercase and underscore characters are required Table 7 2 GlancePlus Highlighting Characteristics Hitem Default Description CPU_CM 20 CPU usage in compatibility mode percentage of total CPU_SW_CM 500 Switches from compatibility mode to native mode switches second CPU_SW_NM 500 Switches from native mode to compatibility mode switches second CPU_PROC 1 CPU_FOPEN 5 Rate of file opens opens second CPU_DISPQ 10 Number of processes in the dispatcher queue DISC_UTIL 80 Individual disc drive usage percentage of total DISC_SERV 0
83. his in several ways Indicators HP GlancePlus uses indicators that are carefully chosen and arranged to make the product easier to use Where several indicators might be used to represent a system resource only those that are clear reliable and easy to interpret are used Indicators that require an in depth knowledge of the operating system or are unique to a particular operating system s technique are not used if better alternatives are available Reducing the number of redundant indicators makes it easier to notice when one exceeds its normal ranges Filtering Intelligent filtering of data helps you focus on what is important For example by showing only interesting processes it is easier to identify problems than it would be if you had to scan through every process on the system GlancePlus s ability to identify interesting processes those that are starved for a resource or those that are using resources excessively make it possible to see the forest without examining every tree Highlighting Highlighting draws your attention to important indicators Highlights might indicate why a process is interesting For example Ifa process was starved for a resource the WAIT field is highlighted and the name of the resource needed is displayed Highlighting can also identify an indicator that exceeds a level that experts consider normal For example if you are displaying the CPU display and the number of compatibility mode switches is
84. iX Measurement Interface was enabled Scale 0 to 25 processes This shows the rate of launches during the current time interval number of launches per second A launch occurs every time the MPE ixX Dispatcher gives a process control of the CPU This is the System Average Quantum SAQ for the C queue CQ This value represents the current time limit in milliseconds during which a process in C queue can retain use of the CPU The SAQ value which is recalculated continuously by the dispatcher must be a value between the minimum MIN and maximum MAX parameters of the current CQ setting This is the rate that files are being opened during the current interval number of files opened per second The File Open Rate can be used to judge the workload characteristics of the system A sustained rate greater than 35 files opened per second might indicate a CPU bottleneck condition or possibly poor application design 5 42 Screen Metrics and Statistics Top CPU Consumer In a single line near the bottom of the CPU Detail screen GlancePlus names the process that is consuming the most CPU resources Figure 5 19 CPU Detail Screen Top CPU Consumer The percentage represents the amount of CPU resource that the process used during the last interval The screen lists either the program name or the MPE iX Command Interpreter CI command last issued If an MPE iX system process is the top consumer it is identified by the program name or
85. if the last interval was typical relative to the life of the process Miscellaneous Process Information Below the State and Last Interval bars the screen shows additional information for the process being monitored Wait READY QUEUE 34 Compatibility Mode 7 o o o oo lt 19 Queue Priority D224 Native Mode Switches N NE sec Scheduling State Ready Comp Mode Switches ES lt 68 sec Transactions min a Response Time ccooooooo lt B 0B Figure 5 9 Process Screen Miscellaneous Process Information Additional information includes the following Wait Upon first entering the Process screen this field indicates why a process is currently waiting if the process is active it will be blank For every succeeding interval it shows both the percentage and wait reason for the highest percentage wait reason during that interval While the STATE and LAST bars group wait reasons for better representation the Wait field uses the detailed reasons in appendix C to increase its ability to pinpoint a process slowdown Queue Priority This field shows the execution queue where A AS B BS C CS D DS E ES and the absolute priority number 1 255 for the process at the time it was displayed The priority will display a zero if the process has just terminated Screen Metrics and Statistics 5 17 Scheduling State Transactions min Compatibility Mode Native Mode Switches Comp Mode Switches
86. ific 5 19 fuzzy logic 7 17 G GlancePlus installation 2 1 GlancePlus Pak for MPE ix 1 2 GLANCNFG file 5 27 7 4 7 16 global bars 3 3 7 2 Global command 3 10 Global key 3 9 Global screen 3 9 5 2 6 color emphasized metrics 5 3 highlighted metrics 5 3 interesting processes 5 2 5 3 5 4 underlined metrics 5 3 group access 2 2 capabilities 2 2 guidelines 6 2 H hardware requirements 2 1 Help command 3 9 3 10 3 13 6 2 Help facility 3 13 help online 1 3 highlighted metrics 5 3 6 1 6 2 7 16 hitem 7 16 Index 4 individual disc utilization 5 45 installation GlancePlus 2 1 group access 2 2 group capabilities 2 2 program files 2 2 INTERACTIVE parameter D 5 interesting applications 5 34 bars 5 35 fields 5 35 interesting processes 3 8 3 15 5 2 color emphasized metrics 5 3 components 5 4 5 11 Global screen 5 4 highlighted metrics 5 3 Job Session screen 5 10 5 11 thresholds 7 5 7 7 underlined metrics 5 3 Interesting Process Thresholds Options screen 3 15 Interrupt Control Stack 5 38 I O by Physical Disc command 3 10 itemid 7 18 J job selection 4 1 Job Session screen 5 7 12 interesting processes 5 10 job session bars 5 8 summary fields 5 8 K key words 3 13 L labels 3 9 LaserRX MPE 1 2 5 27 list commands key 3 9 List Users command 4 3 low level analysis 7 17 memory bar 7 2 7 3 components 3 6 Memory Detail command 3 10 M
87. igh level analysis you still want more information you can respond by pressing X again This will give you the Dump Everything level analysis Usually high level analysis is only done on areas of system performance that the Expert facility determined to be a problem Processes will not be examined for disc problems unless there was an overall disc problem indicated In the Dump Everything level analysis every possible type of problem is examined to determine even a remote possibility that a process is causing a problem This can be useful if you are interested in potential database locking problems on your system but the problems are not affecting enough of the system to be considered serious 6 6 The Expert Facility Recommendations For Using The Expert Facility Using the Expert facility will add a small amount of overhead to the GlancePlus program On most systems this additional overhead is about 0 2 CPU it should not cause any problems On systems with many processes the GlancePlus overhead might be higher In this case you can eliminate almost all the Expert facility s additional overhead by using the V command to set XLEVEL LOW instead of HIGH The high level analysis and suggested actions will not be available to you with this setting You can use XLEVEL LOW as a default and change it to HIGH only when you need more detailed information See chapter 7 for details on how to select the default settings for GlancePlus
88. ition Files INTERACTIVE The INTERACTIVE filename parameter acts the same as the FILE parameter except it signifies that only interactive executions of the program sessions are included in the application class BATCH The BATCH filename parameter acts the same as the FILE parameter except it signifies that only batch executions of a program jobs are included in the application class USER The USER job sessionname username acctname group parameter identifies the job or session user account and group to be included in the application class m job session name The job session name is an optional part of a user logon that can be used for identification If this parameter is included it must be terminated by a comma If this parameter is not supplied all job session names are matched m username The username specifies the name of the user as defined by the MPE accounting structure m acctname The accountname specifies the name of the logon account as defined by the MPE accounting structure m group The groupname specifies the logon group of a particular job or session It is optional but if it is included it must be preceded immediately by a comma If this parameter is not supplied all logon group names are matched Application Definition Files D 5 Note Since commas are allowed within the parameter string in the USER parameter they cannot be used to separate one logon string from another You m
89. k The HP GlancePlus Global Screen The Global screen figure 3 1 is the first screen that appears when you start GlancePlus This screen gives you much information in an easy to read display It provides a global systemwide summary of how resources are being used by active processes HPB1787 B 04 00 HP GlancePlus 16 21 56 Global Ma current avg high CPU MI IBP Po cc 52 66 89 Disc l Kec cece cece cece ence rete ee ee eee ence ee eeeeeeeenee i 6 14 29 Mem M MC ES SD DE 100 100 1007 JSNo Dev Logon Pin Program Pri CPUZ Disc Trn Resp Wait S20 216 TUYLA EVAL 16 MNTMENU C152 1 8 0 1 11 0 2 TERM 39 131 DEBRA EVAL 288 MNTMENU C152 1 3 0 5 6 0 3 TERM S45 310 BECKY EVAL ZZ1 MNTMENU C152 1 6 0 1 11 0 1 TERM 52 124 BETHANNE EVAL 233 MNTMENU C152 1 8 0 4 6 0 3 TERM S60 220 EARL EVAL 234 MNTMENU C152 1 4 0 4 12 0 2 TERM s83 315 FRANCES EVAL 6 MNTMENU C152 1 6 a0 Y 0 1 TERM 119 101 AL ACCTG 173 APPRO1 C152 5 77 0 7 15 0 5 TERM S15 311 SALLIE EVAL 321 MNTMENU C152 1 2 0 0 11 0 4 TERM S75 227 VAN ACCTG 212 APPRO1 C152 3 17 0 0 11 0 3 TERM 549 304 DEBBIE EVAL 264 MNTMENU C152 1 6 0 0 14 0 1 TERM 138 121 FRIEDA EVAL 7S MNTMENU C152 2 2 0 7 12 0 6 TERM 142 225 HARBISON EVAL 262 MNTMENU C152 3 17 e 1 aJ 0 2 TERM S55 221 JANIE EVAL 257 MNTMENU C152 2 87 1 6 9 1 0 TERM SZ9 110 VICKY EVAL 291 MNTMENU C152 1 17 0 0 8 0 2 TERM 169 338 JOE TELESUP 136 GLANCEXL B100 2 67 0 0 2 0 0 Global Select Print MPE Time Refr
90. l help If you request a list of commands or help specific to the current screen you will be returned to GlancePlus when you finish that topic and press Return Select general help if you want to get help on more than one topic A menu will be displayed followed by a greater than gt prompt Then you can enter any of the following m Any menu item For example you can enter XPERT to find out about the GlancePlus Expert facility a A menu item and a key word For example you can enter PROCESS STATE to produce a listing of the codes that may appear in the process STATE bar m Menu to display the Help facility menu m Help to display the Help facility instructions a Exit or E to terminate the Help Facility If you access Help and the softkey feature of GlancePlus is enabled a base set of softkeys will be loaded These base keys will allow you to move directly to Help menus for the four most popular screens or to access additional general Help commands You can access the following m Global screen Help m Job screen Help m Session screen Help m Process screen Help m GlancePlus commands m Main Help menu n Help on Help m Exit Help and return to GlancePlus processing Getting Started 3 13 If you press while in the Help Menu Help displays information up to the next key word or command For example after you enter SESSION Help displays the overall format of the Session screen and a list of key words ass
91. l screen a Memory Detail screen Screen Metrics and Statistics 5 1 Global Screen This screen summarizes activity systemwide and lists all processes that exceed the usage thresholds for the system figure 5 1 The Global screen is the first screen to appear when you start GlancePlus unless you have specified an INFO string in the RUN command to call up another screen see chapter 7 This Global screen is the usual starting point for a review of system activity and performance The information on this screen will tell you if a resource is used excessively or a process is monopolizing available resources HPB1787 B 04 00 HP GlancePlus 14 44 46 Global Ma current avg high CPU Ms SI X ERETTE 87 S4 J17 Disc MSL Mee cece cc cee cee ee cette tees eee eeeee i 12 9 31 Mem M MC ES SD DE 100 100 1007 JSNo Dev Logon Pin Program Pri CPUZ Disc Trn Resp Wait IXPERT e S GR S203 134 SHARON EVAL 182 MNTMENU 2 827 0 0 E 0 2 TERM S211 310 BECKY EVAL 81 MNTMENU 1 87 0 0 Mi 0 4 TERM S5209 23 COLLECT EVAL 212 REVALINQ 3 57 0 6 El 0 4 TERM SZ30 314 MARYANN EVAL 291 MNTMENU 2 124 0 1 E 0 3 TERM S5288 215 VIRGINIA EVAL 228 MNTMENU 2 124 0 1 i 0 1 TERM S208 227 VAN ACCTG 288 APPRB1 0 37 0 0 Z Y TERM SZ93 332 JOE TELESUP Z240 GLANCEXL 3 67 0 0 4 B 6 5295 334 DUP 13 DIALUP 312 CI 0 77 0 2 2 0 0 5295 334 DUP 13 DIALUP 237 LOGON 1 87 0 0 Z 2 5 5295 334 DUP 13 DIALUP 48 RESMNT 9 67 0 5 Z 0 4 5295 334 DUP 13 DIALUP
92. ld to 8 For keywords with two value choices such as NEW YES no simply typing the keyword such as NEW or N will toggle the value to the other choice The capitalized value is the current selection 3 16 Getting Started Different situations will require different thresholds and every system is unique It is worthwhile taking time to become familiar with the Process Thresholds Definition screen and to determine which thresholds provide you with the best information on your system Once you have determined the optimal settings for your thresholds you can have GlancePlus default to your custom values every time you run it see chapter 7 Getting Started 3 17 Commands Additional Information Overview This chapter provides additional information to help you understand how certain HP GlancePlus commands function Selecting Jobs Sessions and Processes gt J S P There are three methods to select the job session or process you want to monitor The first method uses the Next Logical Display command character gt to zoom in with a single keystroke on the process that consumed the most resource during the last interval You cannot use it to select a job or session For example if you are displaying the Global or CPU Detail screen this command will take you to the Process screen of the highest CPU consumer If you are displaying the Filter screen you will go to the Process screen of the highest CPU consumer in the
93. le This line shows the time the CPU was idle during the current interval Queue Distribution The Queue line shows the active CPU cycles consumed during the current interval distributed by queue The percentage for each queue is shown at the right Figure 5 17 CPU Detail Screen Queue Distribution Miscellaneous Performance Metrics The bottom section of the CPU Detail screen shows various Global CPU performance metrics measured during the current interval Switches To CM S Lon o lt 190 sec Current Ready Queue B lt Z Switches To NM S lt 1965 sec Maximum Ready Queue D D lt 21 Interval CM7 lt 237 Launch Rate 131 sec CQ SAQ 19 Interval Process Completions 0 13 sec Interval File Open Rate 26 6 sec Figure 5 18 CPU Detail Screen Miscellaneous Metrics Screen Metrics and Statistics 5 39 Switches to CM This shows the rate of switching from Native Mode to Compatibility Mode Bar Scale 0 to 500 switches per second The number to the right of the bar shows the actual switch rate per second up to a maximum of 9 999 Switches to NM This shows the rate of switching from Compatibility Mode to Native Mode Bar Scale 0 to 500 switches per second The number to the right of the bar shows the actual switch rate per second up to a maximum of 9 999 5 40 Screen Metrics and Statistics Note Excessive switching may indicate CPU saturation While switching rates depend on machine
94. lication definition a process will not belong to that application unless a match for that category is made If no parameters in a category are used in an application definition then a process does not have to match that category For example there can be any number of FILE and USER parameters following an APPLICATION directive but if there is no QUEUE parameter a process could belong to any queue and still belong to the application It would have to match at least one of the FILE parameters and one of the USER parameters Or as another example if an application consisted of USER and QUEUE parameters but had no FILE INTERACTIVE or BATCH parameters then any program file name could belong to the application as long as the user logon and execution queue matched one of the USER and QUEUE specifications Use the OR parameter when you want more than one application definition to apply to the same application You can define up to 15 applications on MPE V and 31 applications on MPE iX GlancePlus predefines an application called OTHER that collects all processes not defined by APPLICATION directives in the application definition file If a process file is included in more than one application it is logged in the first application in which it is defined APPLICATION Order Processing FILE 0FICPROD PAPRPROD SUPPLIES APPLICATION 0ffice Products FILE FURNITUR OFICPROD ELECTRON OFICPROD Application Definition Files D 3
95. m library Library CM System Compatibility Mode system library Library 5 48 Screen Metrics and Statistics Miscellaneous The Miscellaneous section presents several system memory performance metrics that could be used to indicate potential memory bottlenecks A bottleneck occurs when a memory resource has reached its capacity limit and might slow system throughput or response time MemMgr Mi o o oooonoonoo ooo eee lt 4 Page Fault Rate E Po lt 25 2 sec Page Overlay Candidate Rate 5 4 sec System Library Page Faults B 1 sec Memory Mgr Clock Cycle Rate 7 Physical Memory Size 256 mb Figure 5 25 Memory Detail Screen Miscellaneous Metrics MemMgr Memory Manager utilization during the current time interval is shown on the bar and by percentage to the right of this field Sustained utilization of 8 or greater may indicate a potential memory bottleneck Page Overlay This metric depicted as the number of potential free pages Candidate Rate per second shows the Memory Manager activity on the making of physical memory pages as Overlay candidates potential free pages A sustained rate greater than 35 free pages per second may indicate a memory bottleneck Screen Metrics and Statistics 5 49 Memory Mgr Clock Cycle Rate Page Fault Rate System Library Page Faults Physical Memory Size The bottom line of the Memory Detail screen shows the memory manager clock cycle rate The clock cycle rate is computed by dividing
96. m processes by entering the A command Interesting Process Components Each process related to the job or session is displayed in a single line The following components make up a line of interesting process data PIN PROGRAM PRI CPU DISC PIN is the process identification number for the process PROGRAM refers to the name of the program being run If the program is the MPE Command Interpreter CI then this field displays a colon followed by the first characters of the last CI command issued PRI is the execution queue where A AS B BS C CS D DS E ES followed by the absolute execution priority number of the process 1 255 If a process has just terminated then this field will contain the word DIED CPU is the percentage of the central processing unit used by this process during the last interval DISC is the disc transfer rate transfers per second for this process during the last interval Screen Metrics and Statistics 5 11 TRN RESP WAIT ELAPSED CPU TOT DISC TOT P PIN TRN is the rate at which the Return or Enter key was pressed during the last interval A rate is displayed to maintain independence from the time interval thus one transaction in a 30 second interval equates to a rate of 2 transactions per minute The number of transactions is 0 for most batch jobs because generally they do not read from terminals RESP is the average time in seconds between when the Retu
97. manual was issued Many product updates and fixes do not require manual changes and conversely manual corrections may be done without accompanying product changes Therefore do not expect a one to one correspondence between product updates and manual updates Edition 1 6 90 B1787 90001 E0690 Edition 2 4 92 B1787 90001 E0492 Conventions Capitalization Computer Font Capitalized first letters signify HP GlancePlus commands screen names screen sections function keys company names and product names Computer font represents screen text including prompts and messages that appear on the screen and characters or words that you are prompted to enter on the screen Italics Italics are used to emphasize words phrases or characters in the text or in syntax strings to indicate variables UPPERCASE Words in all uppercase capitals represent program fields and system commands that should be entered exactly as shown Return or Depending on your keyboard one or the other represents the terminal key used to execute a command Notice MPE iX Multiprogramming Executive with Integrated POSIX is the latest in a series of forward compatible operating systems for the HP 3000 line of Hewlett Packard computers MPE iX is a superset of its direct predecessor MPE XL and all programs written for MPE XL will run under MPE iX without change Also you can continue to use MPE XL system documentation although it may not refer to f
98. minal response time in seconds during the last interval WAIT This field indicates the current resource for which the process is waiting Screen Metrics and Statistics 5 33 Workload Screen The Workload screen summarizes activity for user defined collections of processes called applications You can use it to view information on resource usage and transactions for groups of specific processes Application definitions are specified in a configuration file for more information see appendix D The detail area of the display is divided into two parts m Interesting Applications a Top CPU Consumer and Top Disc Consumer HPB1787 B 4 HP GlancePlus 13 01 48 Workload Display Ma current avg high CPU MST ip BP x Poe kee eee eee 34 49 76 Disc MMSI 18 Mo cc i 38 44 101 Mem ff MC ES SD DE 100 100 1007 Application CPU current DISC current Trn Resp OTHER n ce ccc cee eee eee lt o Aas rar lt 3 a 0 8 Networking _ _ __ca cece cece rece eccece lt A lt B 3 0 3 Corp Utility oooo m oo moo ooo oo lt BGK Reece cece cee ee ee eee lt 2 a 0 8 Fred A lt 11 E RUU lt 26 17 0 1 Promis E roo Pm o ooo lt TA Re ee ee eee eee lt 3 22 0 9g System Support Sr O gt A lt 2 4 0 6 Top CPU Consumer is Pin 1529 at 12 3 running program RPSZB00P RPSPRG PRODSYS Top DISC Consumer is Pin 1529 at 15 8 1 0 s per sec RPSZBB0P RPSPRG PRODSYS Global Select
99. mp ME ES SD DE 3 100 100 1007 Figure 3 3 Global Bars Getting Started 3 3 CPU Bar HI IBP ee IN S2 66 89 Fa i 6 14 29g ME ES SD DE 100 100 1007 The CPU bar shows how much of the central processing unit s time was spent on various activities during the last interval The current average and highest percentages are displayed numerically to the right of the bar The length of the bar indicates the percentage of CPU time used or time that could be used if pending disc I Os were satisfied during the current interval Paused for disc is not included in these numbers since the CPU is inactive during this time because of pending disc transfers Scale 0 to 100 percent The components of the CPU bar are as follows M MPE iX functions such as memory management interrupt handling and deciding which process uses the CPU System processes such as data communication monitors and spoolers Interactive session processes Batch job processes Paused for disc no CPU is used but Disc I O is occurring Average usage of this resource during this execution of GlancePlus not including Paused for Disc Note The CPU bar may be configured to show activity by queue which changes the above components Chapter 7 contains information on customizing the GlancePlus global bars 3 4 Getting Started Disc Bar HI IBP ee IN S2 66 89 Fa i 6 14 29g ME ES SD DE 100 100 1007 The Disc bar shows the rate of
100. n t have to be clearly true or false but might be a little bit true or very much true Basically the rules consist of comparisons between two items with an accompanying probability value If the rule is true the probability that the symptom is also true increases by the rule s probability value Items are formed either by using constants or by selecting them from a list of available performance indicators supplied by GlancePlus at each interval Items may also be combined into user defined variables The syntax for this section of the GLANCNFG file is described below For an example of the CPU bottleneck rules see below or examine the GLANCNFG file VAR A TOTAL CPU BATCH CPU INTERACTIVE SYMPTOM GLOBAL CPU BOTTLENECK TYPE CPU RULE VAR A gt 70 PROBABILITY 25 RULE VAR A gt 80 PROBABILITY 25 RULE VAR A gt 90 PROBABILITY 25 RULE CPU QUEUE gt 3 PROBABILITY 75 Customizing HP GlancePlus 7 17 To define a user variable express the following VAR A wee Z itemid itemid mnemonic Twenty six variables can be defined VAR A through VAR Z The definition of one variable may include another item but only if that variable has a lesser name alphabetically For example VAR B can include VAR A but not VAR C The mnemonic is the name used to represent this variable when the Expert facility prints out reasons to the user To define a symptom express the following SYMPTOM symptom name TYPE symptomty
101. nd center columns are thresholds that determine which processes are interesting Current Threshold Values System ID HP3 SERIES 932 APPL Defn GLANCNFG PUB SYS 8 Applications defined MinCPU 1 07 WaitIMPEDE 58 0 XPERT CONTINUE demand CPU 10 07 waitCPU 50 07 XLEVEL LOW high DISC 10 sec waitDISC 50 0 RESP 5 sec vaitMEM 50 07 FILEDISP simple DETAIL TRANS 10 min PROCKEYS continue DEMAND NEW YES no DIED YES no DISPCPU INTER BATCH queue DISPMEMORY CONTENTS faults CHANGE lt keyWword newvalue gt Figure 3 7 Setting Process Thresholds In this example GlancePlus displays any process that uses more than 1 0 of the CPU during the current interval and will display and highlight any process that uses more than 10 CPU more than 10 physical disc transfers per second receives an average response time larger than 5 seconds performs more than 10 terminal transactions that is pressing Return or Enter per minute is newly created or dies terminates during the current interval The center column of thresholds causes GlancePlus to display and highlight any process that waited longer than 50 of the interval time for impedes CPU disc or memory To change a threshold type in its keyword followed by a space or an equal sign and the new value To save keystrokes you need only type in enough characters to make the keyword unique Typing either C 8 or CPU 8 would set the CPU thresho
102. ng to be reawakened by its son process When one process creates a second process the first process is referred to as the father and the second as the son The father process has the capability of starting the son and continuing to process or starting the son and waiting until it is reawakened by the son The process is waiting for the completion of a terminal read B 4 Wait Reasons Expanded Definitions Detailed WAIT Reasons Table C 1 Detailed WAIT Reasons Reason Description Reason Description NM Code NM Code Fault PortWait Port Wait WM Stack NM Stack Fault MailWait Mail Wait NM Trans NM Trans Fault JunkWait Junk Wait FPgFault File Page Fault Message Message Wait CM Code CM Code Fault Impeded Impeded CM Stack CM Stack Fault Break Wait for Break CM Trans CM Trans Fault Queue Wait for Queue TermRead Terminal Read Wait Mem Mgmt Wait for Memory Management TermWrit Terminal Write Wait PortMake Port Blocked Make Present Disc IO Disc I O Wait Blocked Wait for Blocked File Other IO Other I O Wait Unblocked Wait for Unblocked File Preemptd Preempted CPU Stor Mgr Wait for Storage Manager SIR Wait SIR Debuglls g Wait for Debug Message to User RIN Wait RIN Wait I0 Confg Wait for I O Configuration MemMgrPF Memory Manager Prefetch PFPReply Wait for port facility process PFP QuantExp Quantum Expired Reply Paused Timer Wait DB Mon Wait for data base DB Monitor Father Wait for Parent DiscFill Wait for Dis
103. nging to this Job Jnnnn or Session Session Processes Snnnn are considered interesting here for the same reasons as in the Global screen see Interesting Processes under Global Screen earlier in this chapter Any job or session process that is not waiting for another process father son or message waited will also be displayed For example the screen will display the command interpreter CI process for a job or session at the top of the Interesting Processes section of the screen If a son process starts running both processes will be displayed while both are running If the CI process must wait for the son process to complete the CI will disappear from the screen during the wait interval but it will return when the wait period ends 5 10 Screen Metrics and Statistics This allows you to see the most interesting part of a job or session without having to manually track the process numbers as they are created and terminated You can see all processes belonging to the job or session by entering the A command Interesting System Use the Select command s SYS option to display only system Processes processes that do not belong to a specific job or session Since there are usually many of these processes left in different waiting states system processes will only be displayed if they are interesting see Interesting Processes under Global Screen earlier in this chapter You can see all syste
104. nteresting 3 8 3 15 5 2 5 10 7 5 7 7 selection 4 1 setting thresholds 3 15 statistics 3 8 thresholds 7 5 7 7 process keys parameter 7 11 Process screen 5 13 25 function keys 5 19 miscellaneous process information 5 17 process bars 5 14 process state bars 5 16 process summary fields 5 15 process specific functions 5 19 Process Thresholds Options screen 3 15 prockeys parameter 5 19 product name 3 3 version 3 3 program access 2 2 Q queue 7 3 changing 4 3 circular 5 35 circular priority 4 3 linear 5 35 linear priority 4 3 priority 5 17 5 35 7 3 QUEUE parameter D 6 QZAP parameter 7 14 QZAPSYS parameter 7 14 R rate file open 5 42 Refresh Screen command 3 10 4 2 refresh time 7 1 parameters 7 15 requirements hardware 2 1 Index 6 software 2 1 resetting statistics 3 7 resource statistics 3 8 response time components 5 16 parameter 7 8 RESTORE group capabilities 2 2 rules defining 7 19 RUN command 3 1 parameters 7 1 RXForecast MPE 1 2 S scheduling state 5 17 screens CPU detail 5 37 disc detail 5 44 filter display 5 29 filter select 3 8 global 3 9 5 2 6 Global 3 2 identifier 3 3 job session 5 7 memory detail 5 47 printing 3 14 process 5 13 process thresholds options 3 15 sections 3 3 selection 7 1 set option values 3 8 workload 5 34 screen sections banner line 3 3 detail display 3 3 3 8 fault types 5 48 filte
105. o compatibility mode switches second NM SWITCHES Rate of switches from compatibility mode to native mode switches second TOTAL MIPS Relative processor speed millions of internal clock cycles per second 7 20 Customizing HP GlancePlus A Multiprocessor Systems In addition to single processor systems MPE now supports multiprocessor systems systems that have more than one processor CPU board To determine if you have a multiprocessor system examine the banner line of the GlancePlus Global screen If more than one processor is configured the total number of configured processors will be displayed in the area to the right of the word Global If this area is blank your system has only one processor configured The following is an overview of how GlancePlus presents CPU metrics on a multiprocessor system Global CPU bars and their associated percentages are presented relative to all processors on the system Any process CPU metric is presented relative to a single processor For example if a two processor system has only one active process that constantly consumes CPU then the Interesting Process section of the Global screen will show 100 CPU for that process The Global CPU bar will display 50 100 divided by the number of processors If that same system had two processes constantly consuming the CPU then each process would show 100 CPU and the Global CPU bar would display 100 100 plus 100 divided by th
106. o display and highlight the data that is most important to you and your system You won t have to wade through unnecessary data when dealing with a problem For example you can easily set thresholds so the Global activity screen displays only processes that exceed your defined limits HP GlancePlus provides data at the level of detail you need You can go from displaying simple graphical global data to displaying technical data to displaying detailed process data By letting you zoom in on the data you need you won t be overwhelmed with too much or too little data at one time HP GlancePlus provides an extensive online help facility developed by performance experts Any help you might need to use the product is right at your fingertips Introduction 1 1 How You Can Use HP GlancePlus You can use HP GlancePlus a As a diagnostic tool to identify immediate performance problems You can easily find out who or what is slowing your system and identify particular programs or applications that are CPU or I O hogs m As a monitoring tool to learn about your system By routinely checking system activity you can document typical performance levels and characterize typical system loads This can make it much easier to identify unusual performance or potential problems Having this knowledge of typical system performance can even help you avoid potential problems with your systems Will You Need Other Performance Tools
107. ociated with it CPU DISC TOTALS PROCESS If you press Return Help displays the CPU topic for Session If you press Return again after the prompt Help displays the Disc information for Session and so on This is similar to turning pages in a manual Printing a Screen The Print facility provides a default formal designator for the print command the file is GLANCELP The default device class is LP Once enabled the Print facility prints each screen as it is updated To print press Print The softkey changes to Print Enabled And the following message appears Printing is now enabled When printing is enabled a hard copy will be printed each time the screen updates and will continue to be printed if you select a new screen To disable the Print Facility press Print Fnabled The following message appears Printing is now disabled The spool file is released to print 3 14 Getting Started To change the default device m Use the MPE command to enter a file equation FILE GLANCELP DEV device The default option CCTL will print one screen per page To save paper by printing contiguously m Add the NOCCTL option to the file equation FILE GLANCELP DEV LP NOCCTL Setting Process Thresholds Once you are familiar with GlancePlus you may want to customize the Global screen to focus attention on the data most important to you and your system environment For example if a 4 sec
108. oggles continuous printing on or off for the line printer LP Change Q Changes queue or priority of a given job Queue session or process Refresh R F Clears and displays an updated screen Screen Time QrQ E5 Changes refresh time or interval Adjust Users v Lists users of a file Set Option v Displays current option values and allows Values you to change them Xpert x Displays Expert Analysis Analysis Zero E Resets averages to zero and recomputes Averages them beginning with the current interval MPE B F4 Allows you to enter an MPE command Command Ctr Terminates command display Ctrl H S Stops scrolling of screen display Ctr Q Resumes scrolling of screen display Update Enter or Updates display immediately Display Space Bar Getting Started 3 11 Online Expert Assistance for Understanding Performance Data Now that you re familiar with the Global screen and have an idea of how to display performance statistics you may wonder how you can get help understanding what the data means All you need to do is use the Expert facility in GlancePlus It provides expert assistance online by a comprehensive set of rules developed by performance experts to alert you whenever a possible performance problem arises You can set an option to get expert assistance continuously or only when you request it using by typing the X command Refer to chapter 6 for details on the Expert facility HPB1787 B
109. ogram BREAK CONS Waiting for reply to a console request CPU Waiting for CPU resources DBMS Waiting for SQL monitor DISC Waiting for disc I O FATH Waiting for father IMP Waiting impeded I O Waiting for nondisc nonterminal read I O MEM Waiting for memory swap MISC Waiting for miscellaneous reason MSG Waiting for user or system message OTHR Waiting for other unknown reasons PAUS Waiting for requested time period to expire RIN Waiting for shared resource such as a file lock SEM Waiting for software semaphore SIR Waiting for critical system resource SON Waiting for son TERM Waiting for terminal read to complete 5 6 Screen Metrics and Statistics Job Session Screen If you select a specific job or session using the J or S commands the Job Session screen appears containing information about how the job or session is utilizing system resources The job or session number is displayed in the Banner Line as part of the screen identifier You may also select all system processes that do not belong to any job or session by using the command and specifying SYS at the prompt In this case System Processes will display as the screen identifier The Job Session screen can be used to track the resource usage of several programs that execute in a job without having to change screens It is shown below The detailed data area of the display is divided into two parts a Job Session Bars and Summary Fields m Interesting
110. ond response time is acceptable and a 5 second response time is unacceptable you would want to see processes receiving 5 second or longer response times highlighted on the Global screen Processes exceeding a given level or threshold as in the example above are called interesting A process can be interesting for these reasons m It uses a large amount of system resources a It must wait an exceptionally long time to use a system resource m It is newly created a It was terminated during the last interval These interesting processes are displayed on the Global screen Some of these interesting processes might be highlighted Highlighting is used to indicate a process that exceeds some higher threshold value For example you may want all processes exceeding 1 of the CPU displayed and processes using more than 10 of the CPU highlighted On color monitors highlighting is in color otherwise highlighted items are underlined The power of GlancePlus lies in its ability to focus your attention on the pertinent data A good rule of thumb is to set your process thresholds so that Getting Started 3 15 the number of interesting processes does not exceed the capacity of a single screen for an interval You also want to set them appropriate to your system environment Use the Set Option Values V command to set process thresholds Figure 3 6 shows the display resulting from the Set Option Values V command The values in the left hand a
111. or session to or from the B subqueue BS At the Enter new queue prompt an SM user can enter a specific number to set the priority at any number between 100 and 255 or enter B to set the priority at 100 The system manager can limit the use of this command to certain user capabilities by modifying the GlancePlus configuration file See chapter 7 for information about customizing GlancePlus The priority of a process or process related job session in a circular queue is placed initially at the base of the chosen queue and will vary within the queue limits The priority of a process in a linear queue remains constant Furthermore a linear queue process will not relinquish the CPU for processes with lower priorities unless it must wait for another resource Caution Misuse of this command can degrade system operating efficiency significantly List Users of a File U Use this command to check the current file users on the system Press U to initiate an on screen request for the FILE GROUP ACCOUNT information for a particular MPE file Based on that information all users of the specified file are listed by process identification number PIN user logon job or session number and logical device number Ldev For example this can be used to identify all users of a program file you want to replace Note MPE message files are not supported because of the special type of control blocks that govern access to them Commands
112. pe where symptomtype can be one of the following CPU DISC MEMORY IMPEDE SWITCH CM RESPONSE or OTHER One or more rules defining when the symptom occurs should follow each symptom The syntax for defining a rule is RULE itemid comparison itemid PROBABILITY percent where itemid can be a constant a user variable or one of the GlancePlus performance indicator mnemonics gt greater than lt lesser than gt greater than or equal to lt lesser than or equal to Percent is a value from 100 to 100 Comparison is You may define as many as 50 symptoms and 100 rules at one time GlancePlus performance indicators are defined below 7 18 Customizing HP GlancePlus Indicator TOTAL CPU SYSTEM CPU SESSION CPU BATCH CPU MEM MGR CPU DISPATCH CPU OTHER CPU CPU PAUSED IDLE CPU PHYS DISC SYSTEM DISC SESSION DISC BATCH DISC MEM MGR DISC LOGICAL DISC DISC UTIL TRANSACT MIN FIRST RESP RESPONSE Table 7 3 GlancePlus Performance Indicators Description Total CPU usage in the interval percentage of available CPU usage by system processes percentage of total CPU usage by interactive session processes percentage of total CPU usage by batch job processes percentage of total CPU usage to manage main memory percentage of total CPU usage used in dispatching processes percentage of total CPU usage for other reasons percentage of total Time CPU was idle but disc IO occurred percentag
113. program data stacks D Data objects certain system data structures and compatibility mode data segments F Files data files such as TurboIMAGE data sets KSAM key and data files and normal MPE files Average usage of this resource during this execution of GlancePlus A particular category will not be displayed if there is insufficient Memory usage to occupy one character space on the screen 3 6 Getting Started Note The Memory bar may be configured to show fault rates which changes the above components Chapter 7 contains information on customizing the GlancePlus global bars Percentage Columns The percentage columns at the right of the global bars figure 3 3 show current average and highest use of each system resource Current value represents use for the current interval Average is the mean value of all usage data collected since the present monitoring session began High value shows the single highest usage amount of that resource during the monitoring session If you reset the statistics to zero with the Zero Averages command Z the current value will still represent current usage for each interval but subsequent average and high values will represent data collected since the resetting Note Upon entry to GlancePlus the first global screen will show GLANCEXL using a high percentage of CPU This is normal since the program must take two data samples within a few seconds of each other in order to displa
114. r bars 5 30 filter processes 5 32 function keys 3 3 global bars 3 3 individual disc utilization 5 45 interesting applications 5 35 interesting processes 5 2 5 3 5 10 job session bars 5 8 miscellaneous 5 49 miscellaneous performance metrics 5 39 miscellaneous process information 5 17 process bars 5 14 Process screen function keys 5 19 process state bars 5 16 process summary fields 5 15 queue distribution 5 39 summary disc utilization 5 46 summary fields 5 8 5 31 top CPU consumer 5 36 5 43 top disc consumer 5 36 5 46 type of utilization 5 38 Select command 3 10 4 1 Select Job command 3 10 4 1 Select Process command 3 9 3 10 4 1 Select Session command 3 10 4 1 service time 7 16 session selection 4 1 Set Option Values command 3 10 3 16 5 3 5 19 5 20 6 5 6 7 7 4 Set Option Values screen 3 8 setting parameters 7 5 single letter commands 3 2 3 10 softkeys 3 9 software operating system 2 1 Software Performance Tuner SPT 1 2 data collection 5 25 starting GlancePlus 3 1 state bars process 5 16 components 5 16 statistics process 3 8 resetting 3 7 resource 3 8 summary disc utilization 5 46 summary fields 5 31 process screen 5 15 switches 5 20 7 16 between modes 5 18 guidelines 5 39 5 41 5 42 symptoms 7 18 defining 7 19 SYS option 5 10 system processes display SYS 5 10 T terminal requirements 2 1 thresholds 3 15 17 setting 3 15 values 5
115. rn or Enter key was pressed and the computer was ready to accept more input This number is 0 0 if there were no transactions during the last interval WAIT is the reason a process was waiting at the end of the last interval See table 5 1 for definitions of wait reasons and appendix B Wait Reasons Expanded Definitions More detailed Wait reasons are defined for the Process screen see appendix C ELAPSED is the length of time a process was alive CPU TOT is the amount of CPU time used during the life of the process DISC TOT is the total number of physical disc transfers completed during the life of the process A number followed by the letter K represents thousands of transfers a number followed by the letter M represents millions of transfers PIN is the process identification number for the parent or father of this process Note Disc totals are not collected until either HP GlancePlus or another performance tool enables the MPE measurement facility to collect process data This means that the DISC TOT value will not include physical transfers that were completed by a process before the MPE measurement facility was enabled 5 12 Screen Metrics and Statistics Process Screen If you select a specific process using the P or gt commands the Process screen appears containing information about that process The Banner Line includes the PIN number and job or session number for the process as part of
116. s the percentage of time the CPU is busy not counting idle and paused for disc time An asterisk represents the average usage of this resource during this execution of HP GlancePlus You can use the Z command to reset the averages to zero and recompute them beginning with the current interval values This bar shows the job or session s disc transfer rate transfers per second during the interval The codes in this bar correspond to the codes used in the Global Disc bar see Disc under Global Screen earlier in this chapter The overall transfer rate appears at the right of the bar An asterisk represents the average usage of this resource during this execution of HP GlancePlus You can use the Z command to reset the averages to zero and recompute them beginning with the current interval values 5 8 Screen Metrics and Statistics COUNT LOGON ELAPSED CPU TOT DISC TOT The count in parentheses is the number of processes associated with the current job or session There are four processes shown in the previous example LOGON is the logon string jobsessionname user account The logon string in the previous example is DIANNE EVAL ELAPSED is the total clock time during which a job or session has been logged on CPU TOT is the total amount of processor time used during the life of the job or session see note DISC TOT is the total number of physical Disc transfers completed during the lif
117. second GLANCNFG THRESHOLD DISC percent JCW SETJCW GLANCEDISC iopersec VALUES DISC iopersec Customizing HP GlancePlus 7 7 The RESPONSE TIME parameter determines what terminal response time a process must exceed in order to be displayed with its response time field highlighted DEFAULT 5 seconds GLANCNFG THRESHOLD RESP seconds JCW SETJCW GLANCERESP seconds VALUES RESP seconds The TRANSACTIONS parameter determines the terminal transaction rate that a process must exceed in order to be displayed with its transaction rate field highlighted DEFAULT 10 transactions minute GLANCNFG THRESHOLD TRANS transpermin JCW SETJCW GLANCETRANS transpermin VALUES TRANS transpermin The NEW or NONEW parameter determines whether or not processes that are newly created are displayed with their names highlighted DEFAULT New processes are displayed and highlighted GLANCNFG THRESHOLD NONEW Specifies to not highlight new processes JCW SETJCW GLANCENEW 1 Specifies to not highlight new processes SETJCW GLANCENEW 0 Specifies to highlight new processes VALUES NEW YES or NEW NO YES means to highlight new processes 7 8 Customizing HP GlancePlus The DIED or NODIED parameter determines whether or not processes that terminate are displayed with the process priority field highlighted as DIED DEFAULT GLANCNFG JCW VALUES Terminated processes are displayed and highlighted THRESHOLD NODIED Specifies to not hi
118. sion or process to monitor you must follow the command abbreviation for that screen with a valid number Customizing HP GlancePlus 7 1 The following examples specify a refresh time of 60 seconds and the monitoring of job number 312 m Type RUN GLANCEXL PUB SYS INFO J312 PARM 60 a If you activated the HP GlancePlus UDC file type GLANCE J312 60 The following examples specify that startup should proceed directly to the Workload screen m Type RUN GLANCEXL PUB SYS INFO W a If you activated the HP GlancePlus UDC file type GLANCE W Customizing GlancePlus Global Bars As discussed in chapter 3 the Global CPU bar and the Global Memory bar can be configured differently This will change the abbreviations and their corresponding components in those bars For example you can reconfigure the components of the Global CPU bar to show activity by queue and the components of the Global Memory bar to show fault rates as discussed below Global CPU Bar Reconfiguring Global CPU Bar to show activity by queue is useful when you use a batch job to initiate processes that interact with terminal users The normal interactive batch representation would show these terminal users as batch since they were initiated as a batch job this would not be as useful as a queue representation A queue representation might also be useful on a system where interactive sessions and batch jobs routinely execute in queues other than t
119. ss Screens Command Command Function Action Name Character Key CPU Detail Displays CPU Detail screen Disc Detail Displays Disc Detail screen Filter 4 Displays Filter screen Global O F1 Displays Global screen Help H F7 Enters Help facility Select Job 0 Selects particular job to be displayed in detail on the Job Session screen Memory M Displays Memory Detail screen Detail Xpert 0 Displays Expert Overview of system Overview bottlenecks Note The X key is the command character for the Xpert Analysis command Select e Selects particular process to be displayed in Process detail on the Process screen Select s Selects particular session to be displayed in Session detail on the Job Session screen Workload w Displays Workload screen Select 1 12 Selects job session process to monitor Previous Y Displays screen that was last active before Display entering current display Next Logical Displays next logical screen that is the top Display resource consumer See chapter 4 3 10 Getting Started Table 3 2 Commands to Access Program Activities Command Command Function Action Name Character Key All A Displays detail data for all processes or Processes applications rather than only interesting processes or applications Exit Terminates program Glance Print T
120. ters Display control parameters control the actions of various GlancePlus display screens The FILE DETAIL DISPLAY parameter determines whether the display of files opened by a process on the PROCESS screen is in detailed one open file per line or simple four files per line format DEFAULT DETAIL GLANCNFG FILEDISP SIMPLE or FILEDISP DETAIL JCW SETJCW GLANCEFILEDISP 0 for detailed or SETJCW GLANCEFILEDISP 1 for simple VALUES FILEDISP SIMPLE or FILEDISP DETAIL The PROCESS KEYS parameter determines whether or not the OPEN FILES and PROCESS FAMILY commands show only one display on the Process screen each time they are requested on demand or toggle between no display and continuous display continuous DEFAULT DEMAND GLANCNFG PROCKEYS DEMAND or PROCKEYS CONTINUE JCW SETJCW GLANCEPROCKEYS 0 for demand or SETJCW GLANCEPROCKEYS 1 for continue VALUES PROCKEYS CONTINUE or PROCKEYS DEMAND Customizing HP GlancePlus 7 11 The CLEAR SCREEN parameter determines whether old display data is always erased before a new display begins CLEARSCREEN ON or whether some screens scroll previous process data down allowing the current data to be compared against previous data DEFAULT OFF GLANCNFG CLEARSCREEN ON or CLEARSCREEN OFF JCW SETJCW GLANCECLEARSCREEN 0 for off or SETJCW GLANCECLEARSCREEN 1 for on VALUES This parameter cannot be changed using the VALUES command The DISPLAY CPU parameter determines what typ
121. the applications are using CPU and Disc resources and information on their transaction rates and response times The Application bars and fields contain the following information Application This is the name of the application as specified in the Name application definitions see appendix D CPU This CPU bar shows the percentage of central processing unit time used by the application during the last interval The bar contains the following codes m L Linear queue AS or BS m C CS queue m D DS queue m E ES queue The number at the right of the bar is the percentage of time the application used the CPU during the last interval A bar will not be displayed if the percentage of CPU used was not high enough to occupy one character space on the bar Screen Metrics and Statistics 5 35 DISC This bar shows the application s disc transfer rate transfers second during the interval The bar contains the following codes a R Disc reads a Y Disc writes The number at the right of the bar is the application s disc transfer rate during the interval TRN This field indicates the rate of terminal transactions number minute during the last interval RESP This field indicates the average terminal response time seconds during the last interval Top CPU and Disc Consumer GlancePlus names the processes that are consuming the most CPU resources and the most Disc resources on the two lines following the application bars
122. the number of times the memory manager has cycled through memory by the elapsed time GlancePlus has been running Unlike the other metrics on the screen this one is relative to the run time of GlancePlus rather than the current interval This avoids the misrepresentation of reporting one clock cycle over four thirty second intervals as a rate of 120 per second during one interval and 0 per second during the other three It also means that you should be careful to let GlancePlus run a few minutes before drawing any conclusions from this metric A memory manager clock cycle rate of more than 25 per hour over ten minutes or more may indicate possible memory pressure This field shows the sum of the Global System Page Fault types The sum is expressed in page faults per second A fault rate greater than 30 faults per second sustained over several intervals may indicate a potential bottleneck This metric shows the total Native Mode and Compatibility Mode system library faults A sustained rate of greater than 2 faults per second may indicate a memory bottleneck condition This metric indicates the size of the machine s physical memory in megabytes 5 50 Screen Metrics and Statistics 6 The Expert Facility To use a performance management tool effectively you must be able to obtain proper measures of your system s performance then separate the important from the unimportant information HP GlancePlus for MPE iX helps you do t
123. this process during the last interval The DISC field is highlighted if the process exceeds the disc transfer threshold The default is 10 disc transfers per second TRN is the rate per minute at which the Return or key was pressed during the last interval Transactions with think times less than 0 2 second are not counted since they probably represent hardware status replies The number of transactions is 0 for most batch jobs because generally they do not read from terminals The TRN field is highlighted if the process exceeds the terminal transaction threshold The default is 10 transactions per minute RESP is the average time from when the Return or key is pressed until the computer is ready to accept more input This includes the time for the process to do all the work requested by the user write out all the characters in response to the request and issue the next read to the terminal This number is 0 0 if there were no transactions during the last interval The RESP field is highlighted if the response time exceeds the response time threshold The default is 5 0 seconds WAIT is the reason a process was waiting at the end of the last interval For a list of WAIT reasons see table 5 1 and appendix B Wait Reasons Expanded Definitions Screen Metrics and Statistics 5 5 Table 5 1 Wait Reasons Wait Definition BREK Waiting to be resumed from a pr
124. timeout is in effect for all GlancePlus prompts to insure that you do not get an excessive number of interesting processes due to a large time interval when the next screen is displayed If you do not respond to a prompt in 90 seconds GlancePlus will move you to the next screen or prompt There are two exceptions When you are in the Help subsystem or in the midst of executing an MPE command GlancePlus does not have control of the terminal Therefore it cannot initiate a time out It is recommended that you limit your time in these areas to a few minutes 4 4 Commands Additional Information 5 Screen Metrics and Statistics HP GlancePlus for MPE iX provides different levels of performance analysis You can use the statistics on the Global screen to monitor systemwide activity or you can refer to the detailed data screens to focus on specific areas of system usage This chapter describes the metrics and statistics presented on the detailed data screens The discussion for each screen focuses on the Screen Elements display the detailed data area of the screen The Screen Elements display differs for every GlancePlus screen The Banner line and Global bars are the same for every detailed data screen and are discussed in chapter 3 The following detailed data screens are described in this chapter m Global screen m Job Session screen m Process screen m Filter screen m Workload screen a CPU Detail screen m Disc Detai
125. uch of the time the process is requesting service from the computer The components of the State bar are shown below For each component listed the process state was as follows Using the CPU Waiting for memory resources Blocked waiting for a disc I O transfer Blocked waiting for a non disc I O transfer this includes waiting for terminal writes but not terminals reads Remember the State bar represents the portion of time the process is asking for service Q Impeded by a synchronizing mechanism including the following wait reasons DBMS RIN SEM SIR and IMPEDE wait times P Preempted waiting for the CPU HU FBO F Waiting on a process in the family tree including the following wait reasons father son and message wait times 0 Waiting for some other reason Wait reasons in this category include TIMER PORT JUNK MAIL CONSOLE PAUSE TERMINAL MISCELLANEOUS and BREAK wait times 5 16 Screen Metrics and Statistics LAST Components of Response Time The LAST or Last Interval Process State bar represents components of response time for the process over the last interval The number at the right represents the percentage of time the process was asking for service not waiting for input from a terminal during the most recent time interval The codes for the Last bar are similar to those for the State bar except they only represent the time for the last interval A comparison of the LAST bar to the STATE bar shows
126. ust use another character such as a space to separate parameters or use a separate USER line for each logon QUEUE Usually the execution priority of a process is CS DS or ES Special accounts can log on or run programs with PRI BS whereas privileged mode and system processes can place themselves in the AS or linear queues Linear queue can be any priority you want but it is not subject to the normal priority adjustments of the MPE dispatcher You can restrict processes in an application to those belonging to selected execution priority queues using the QUEUE L A B C D E parameter More than one queue can be indicated The following example specifies any process running in either the DS or ES queue QUEUE D E Note The AS and BS queues are special cases of the linear execution queue If you specify QUEUE L then processes executing as AS or BS priority are included Specifying QUEUE A B is not the same as specifying QUEUE L since the linear queue can be outside the range of the A and B queues A process with process handling capabilities can change its execution queue as it runs A process s execution priority can be altered also by an external tool such as HP GlancePlus The process execution queue is sampled at the end of each 1 minute sample interval If the process has changed queue it can change applications All activity for a process during the 1 minute sample interval is assumed to have occurred in the
127. vent backups The Expert Facility 6 7 Changing The Expert Facility s Rules A situation might occur on your system that the Expert facility considers unusual but is normal for your system Or conversely you might have a situation that the Expert facility does not detect but one that you would like to add to its list of symptoms In such cases see chapter 7 for information on changing the GlancePlus default configuration Make changes to the Expert facility carefully since by doing so you are assuming the role of an expert 6 8 The Expert Facility Customizing HP GlancePlus As discussed in chapter 3 GlancePlus allows you to customize the startup procedure Customizing Startup HP GlancePlus for MPE iX provides the following two startup parameters when you run the program Refresh Time Once you execute HP GlancePlus the screens are refreshed updated every 30 seconds by default You can specify a different refresh time when you execute the program by using the PARM option of the MPE RUN command With this parameter you can specify the number of seconds 5 3600 between screen updates Selecting the Initial Screen GlancePlus displays the Global screen at startup by default You can direct it to start with the Job Session Process CPU Detail Disc Detail Memory Detail or Workload screen by specifying the appropriate GlancePlus command in the INFO option of the MPE RUN command To select a job ses
128. will not be redisplayed when the screen is refreshed Screen Metrics and Statistics 5 19 o Continuous When Continuous is selected the functions are treated as switches enabling or disabling the continuous display of information For example if Prockeys Continuous and the Display Files function key is selected from the Process screen the open file information is displayed and redisplayed automatically each time the screen is refreshed When a continuous mode display is enabled the Process State bars and Miscellaneous Process Information sections of the detail display are suppressed to leave more room for the display just enabled When that display is disabled the Process State bars and Miscellaneous Process Information are again included in the detail display Hint If you wish to see successive displays on the monitor for comparison set Prockeys Demand and press the Display Files function key at the prompt Press RETURN to continue or enter a GLANCE command This is useful particularly for tracking the progress of the file pointer in the file display It has the additional benefit of not updating the rest of the process information this can be very handy over a slow data communications link Display Files Press the Display Files function key to see a list of open files for the process being viewed To display this data without using the function keys press 1 The 1 subcommand is only valid during the
129. y the first Global screen At the default 30 second interval CPU usage is typically low and should cause no problems on your system CPU overhead used by GlancePlus is proportional to the frequency of screen update and the number of processes on the system You can cause excessive CPU overhead by pressing a key every few seconds Getting Started 3 7 Detail Display The information presented in the detail display portion of the screen below the global bars depends on the type of screen figure 3 4 JSNo Dev Logon Pin Program Pri CPUZ Disc Trn Resp Wait S20 216 TUYLA EVAL 16 MNTMENU C152 1 8 0 1 11 0 2 TERM 39 131 DEBRA EVAL 288 MNTMENU C152 1 3 0 5 6 0 3 TERM S45 310 BECKY EVAL ZZ1 MNTMENU C152 1 6 0 1 11 0 1 TERM 52 124 BETHANNE EVAL 233 MNTMENU C152 1 8 0 4 6 0 3 TERM S60 220 EARL EVAL 234 MNTMENU C152 1 4 0 4 12 0 2 TERM s83 315 FRANCES EVAL 6 MNTMENU C152 1 6 a0 Y 0 1 TERM 119 101 AL ACCTG 173 APPRO1 C152 5 77 0 7 15 0 5 TERM S15 311 SALLIE EVAL 321 MNTMENU C152 1 2 0 0 11 0 4 TERM S75 227 VAN ACCTG 212 APPRO1 C152 3 17 0 0 11 0 3 TERM 549 304 DEBBIE EVAL 264 MNTMENU C152 1 6 0 0 14 0 1 TERM 138 121 FRIEDA EVAL 7S MNTMENU C152 2 2 0 7 12 0 6 TERM 142 225 HARBISON EVAL 262 MNTMENU C152 3 17 e 1 aJ 0 2 TERM S55 221 JANIE EVAL 257 MNTMENU C152 2 87 1 6 9 1 0 TERM SZ9 110 VICKY EVAL 291 MNTMENU C152 1 17 0 0 8 0 2 TERM 169 338 JOE TELESUP 136 GLANCEXL B100 2 67 0 0 2 0 0 Figure 3 4 An Example of a D

HP GlancePlus User`s Manual

Contents

Download Pdf Manuals

Related Search

Related Contents