Home

Grade of Service - HP Software Support

1. 5 27 LAN MGIricS oi iona aa Si kw a balan Seth and a ea a 5 31 WAN Metris 2440005 naa aiaia a wa aaae a cane beeen ead 5 33 LAN GOS Calculations 00000 eee eee ee 5 36 WAN GOS Calculations aana 00 0 0 0 e ee eee 5 36 Variables Polled from MIB Il ifEntry Table 6 21 Variables Polled from MIB Il System Table 6 23 Variables Polled from MIB II ipAddrEntry Table 6 26 Router Metrics e sarsonu Au paR eee te tees 6 27 Bay Router Variables 0 2 0 0 00 e eee eee 7 26 Bay Router Metrics 0 0 00 cece eee 7 28 Bay Router GOS Calculations 2 0055 7 30 Cisco Router Variables 0 00 0 cee eee 8 30 Cisco Router Metrics 0 2 0 0 00 eee eee 8 34 Cisco Router GOS Calculations 0000 8 36 xiii Tables Xiv 11 1 11 2 11 3 12 1 12 2 12 3 12 4 12 5 12 6 12 7 TREND Database Daily Average and Maximum Variables TREND Database Baseline and Forecast Variables TREND Database Metrics 00 2200055 System Variables Polled from Empire kernelperf Table System Variables Polled from MIB II Server Volume Table System Variables Polled from Empire cpuStatsEntry Table System Variables Polled from Empire diskStatsEntry Table System Variables Polled from Empire devTableEntry Table System Metrics 0 00 cece eee eee System GOS Calculations TREND Preface TREND is collection of r
2. U o i x J9 NOY ODSID 8 8 Cisco Router ReportPack The CPU Utilization pie chart shows the percentage of samples taken during the day in which the CPU Utilization metrics calculated for each sample fall in the defined range CPU Utilization boston Acme s Personnel Office Sun Dee 19 1999 Hi cPu_util_o_20 E cPu_util_20_40 Hi cPu_util_40_so Wi cPu_wutil_so_100 The Free Memory pie chart shows the percentage of samples taken during the day in which the Free Memory Utilization metrics calculated for each sample fall in the defined range Free Memory boston Acme s Personnel Office Sun Dec 19 1999 Hio 19 Utilization 4 20 39 Utilization 40 79 Utilizaiton A Ws0 100 Utilizaiton 8 12 Report Descriptions ReportPack Guide The Total Buffers pie chart shows the percentage of samples taken during the day in which the Total Buffers metrics calculated for each sample fall in the defined range Total Buffers boston Acme s Personnel Office Sun Dec 19 1999 Di butter Util 0 19 Wi buffer Util 20 39 Di Butter Util 40 79 Wi buffer Util 80 100 The Percent Errors line graph shows the average percentage of errors for each hour during the report day For more information about how Errors is computed see Table 7 2 Percent Errors boston Acme s Personnel Office Sun Dec 19 1999 0 00 00 06 00 12 00 18 00 00 00 The Percent Discards line graph shows the average percentage of
3. U D la 2 gt 2 3 2 D 1 If you create your own tab delimited file you must Use the file formats described in Property File Formats on page A 1 Use all columns shown in Property File Formats on page A 1 in the same order Do not use quotes in your file Do not use NULL values in your file Ensure that the property data columns of dsi_target_name and dsi_table_key match the values of dsi_target_name and dsi_table_key in the key table For example if you have a dsi_target_name of router1 witha dsi_table_key value of 1 23 in your property file there must be a matching dsi_target_name of router1 with a dsi_table_key value of 1 23 in the key table Otherwise a new row and keyid value will be inserted into the key table Set the values of the _id values to 2 which is the default value if you are not using some or all of them For example if you are not using region_id set the value for region_idin your file to be 2 Use the following file naming convention lt ReportPackName gt _Property dat Place the property data files in the following folder DPIPE_HOME packages lt ReportPackName gt lt ReportPackName gt ap PropertyData Archive To create your own property file Using the Property Import Feature 1 23 1 ReportPack Administration 1 Create a tab delimited file using an application like Microsoft Excel 2 Edit or add properties 3 Export the prop
4. Warning E Congestion E Out Utilization Win utilization Good Excellent Wed 00 00 Wed 06 00 Wed 12 00 Wed 18 00 Thu 00 00 Figure 3 21 Snapshot Grade of Service Frame Relay Quick View and Snapshot Reports 3 23 swioday wo gt a lt S gt Ko J m z oO 3 Analyzing TREND Reports Hourly Volume Figure 3 22 shows the Hourly Volume area graph which displays the total hourly volume for the selected PVC over the report day Hourly Yolume Sanjose Support Wed Dec 22 12 00 Aw 121 20 90 90 2 m 60 60 oa v 30 30 0 00 00 06 00 12 00 18 00 00 00 E Out volume E in volume Figure 3 22 Quick View Hourly Volume Frame Relay Hourly Congestion Figure 3 23 shows the Hourly Congestion line graph which plots the total number of frames for each hour with Forward Explicit Congestion Notification and the total 3 24 QuickView and Snapshot Reports ReportPack Guide number of frames with Backward Explicit Congestion Notification for the selected PVC for each sample taken during the report day Hourly Congestion Sanjose Support Wed Dec 22 12 00 Aut 34 00 ty Ld in 17 00 bad in Number of Frames 0 00 00 06 00 12 00 18 00 00 00 Frames with FECNS Frames with BECNS Figure 3 23 Quick View Hourly Congestion Distribution of Bandwidth Utilization Input Figure 3 24 shows the Distribution of Bandwidth Utilization Input pie chart w
5. 0 01 Beacons per Second 20 00 00 E Aug Beacons Sec M Peak Beacons Sec 9 RMON ReportPack TREND 9 12 TREND 10 RMON2 ReportPack The RMON2 ReportPack summarizes data collected from RMON probes which provides the network manager with information about network performance o D e Z N D fe U o Q The ReportPack provides four reports Hot Spots QuickView Snapshot and Top Ten that allow you to identify which hosts utilize the majority of your network segment s bandwidth Additionally they highlight slow network response time by showing your segments with high utilization error counts or both This information can help you reorganize your segments based on host traffic patterns Report Descriptions The following sections describe each report in the RMON2 ReportPack in detail 10 1 10 RMON2 ReportPack Hot Spots Report This report lists all network interfaces with utilization greater than 20 and or collisions that amount to greater than 15 of good packets for the preceding day Associated charts provide hourly segment utilization collisions errors and frame size distribution for the selected interface For an in depth discussion of the Hot Spots report see Hot Spots Report on page 3 32 The Hot Spots report comprises the following charts The Problem Summary for the Day table provides a list of interfaces that meet or exceed the specified criteria
6. TREND document and associated file names are as follows Document Name File Name TREND Installation Guide TRENDxx_InstallationGuide pdf TREND User s Guide TRENDxxUsers_Guide pdf TREND ReportPack Guide TRENDxxReportPack_Guide pdf TRENDweb Installation and TRENDwebxx_InstallGuide pdf Administration Guide Guide To Building TREND Reports TRENDwebxx_BuildersGuide pdf Guide To Viewing TREND Reports TRENDwebxx_ViewersGuide pdf where xx is the product version number For example 32 refers to version 3 2 These documents are also installed onto your hard drive in the docs directory in DPIPE_HOME during installation xviii ReportPack Guide You can download the latest TREND manual updates from the DeskTalk Systems Customer FTP site Look at the document number on the manual s title page to determine if you have the latest version of manual IG TREND Installation Guide UG TREND User s Guide RPR TREND ReportPack Guide TW IAG TRENDweb Installation and Administration Guide BTR Guide To Building TREND Reports VTR Guide To Viewing TREND Reports Product version for example 32 means Version 332 Publication date in yyyymmmdd format TW IAG 32 2000MAR06 Call DeskTalk Systems Customer Support for download instructions xix TREND D U D Q 2 gt 2 3 D 1 TREND 1 ReportPack Administration This chapter discusses the following ReportPack dependencies
7. devIndex 1 3 6 1 4 1 546 1 1 1 7 1 1 This object indicates the row instance The instance is derived from the index number from the mounted file systems file devDevice 1 3 6 1 4 1 546 1 1 1 7 1 2 The device name e g dev sd0a corresponding to the mounted file system For example the 7 file system is commonly mounted from the device dev sd0a For network file systems the value corresponds to the remote host and remote file system name e g emptech usr local devMntPt 1 3 6 1 4 1 546 1 1 1 7 1 3 N se 1 of 4 o 1 3 a me fe Data Source 12 47 5 S fo 12 System ReportPack Table 12 5 System Variables Polled from Empire devTableEntry Table Variable OID String The mount point or path name in the local file system where the device is mounted e g home devBsize 1 3 6 1 4 1 546 1 1 1 7 1 4 The block size of the mounted device in bytes File systems and disk devices generally allocate data in fixed size blocks generally referred to as the block size of the device Common block sizes are 1024 1K and 2048 2K bytes devTblks 1 3 6 1 4 1 546 1 1 1 7 1 5 The total number of blocks contained in the mounted device devFblks 1 3 6 1 4 1 546 1 1 1 7 1 6 The total number of free or unallocated blocks in the mounted device devTfiles 1 3 6 1 4 1 546 1 1 1 7 1 7 Each disk device and file system allocates a
8. The ranking table has one row per day for each managed element that is being ranked and the following six columns Timestamp for day1 day2 day3 and so on ID of the metric that is being ranked Daily summary value of the metric being ranked Rank of that metric 1 2 3 and so on Rank for the same metric on the previous day gt gt Delta rank This item represents the absolute value of the difference in the metric s value between yesterday and today A large delta rank indicates a volatile element which you should probably investigate A relatively small delta rank indicates stable element performance Previous delta rank Ranked metrics appear in the Top Ten reports in the TREND ReportPacks Ranking 2 13 2 Overview Frequency Distribution The frequency distribution of a sample is obtained by dividing the sample s frequency in a defined class bucket by the total number of samples For example the In Utilization metric for the Frame Relay ReportPack is calculated for each sample taken during the day for the selected PVC After the In Utilization percentage is recorded a value of is assigned to one of the defined buckets 0 20 21 40 41 60 61 80 81 100 and gt 100 depending on the recorded percentage If the In Utilization percentage for one sample is 42 then a value of 1 is assigned to the 41 60 bucket The values for the buckets are totaled for all samples taken dur
9. tulsa 7 00 0 00 0 00 0 00 7 00 boston 7 00 0 00 0 00 0 00 7 00 Figure 3 30 Hot Spots Problem Summary for the Day Cisco Router Hot Spots Report 3 33 3 Analyzing TREND Reports To help you analyze router problems the summary row for each device is linked to multiple drill down components that bring together the varied information required to investigate potential network trouble spots pinpoint the source of the difficulty and determine possible resolutions For Cisco routers exception conditions are tracked for CPU Utilization Memory Utilization Buffer Utilization and Buffer Faults Reporting these metrics helps you identify routers that are unable to handle the amount of network traffic being generated These indicators tell you when a router needs a CPU upgrade when there is inadequate memory or when buffers are being overutilized The drill down components below the Daily Problem Summary chart supply instant access to detailed information in the areas of Grade of Service CPU Utilization Buffer Utilization Memory Utilization Discards and Buffer Faults for each hour of the last day You can readily see the correlation between critical performance indicators during particular time periods The charts and graphs show exactly when a threshold is breached and what happens before and after the breach Thus you can determine if a problem is a long standing issue that has been growing steadily or a sudden burst that may not
10. 0 Sep 10 Sep 24 OctOs 6 6 Report Descriptions ReportPack Guide The Packet Rate line graph shows the percentage of packets for the selected router for each day of the baseline period Packet Rate sanjose Unassigned Fri Sep 10 1999 Thu Oct 21 1999 28 893 09K 2 20 189 82K a amp D 13 446 54K 4 e 3 6 723 27K p mo o Sep 10 Sep 24 Oct 08 Q a The Total Bytes bar chart shows the percentage of bytes for the selected 2 a router over the baseline period Total Bytes sanjose Unassigned Fri Sep 10 1999 Thu Oct 21 1999 56 086 42 066 28 046 Bytes 14 026 Sep 10 Sep 24 OctOs Report Descriptions 6 7 6 Router ReportPack QuickView and Snapshot Reports The QuickView report provide detailed information on key metrics for individual routers with the highest throughput You can double click a router in the Router Inventory table to display reports that show hourly volume and throughput for the selected router The Snapshot report provides the same information as the Quick View report however it allows you to select a particular router The selected router may be different than the routers appearing in the Quick View report For more information about selecting a router for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The QuickView and Snapshot reports comprise the following components The Rout
11. Average CPU Utilization houston Acme s Support Office Mon Dec 20 12 00 4M Mon Dec 20 11 15 PM 100 00 80 00 60 00 Percent 40 00 20 00 05 00 11 00 17 00 23 00 The Buffer Utilization line graph shows the buffer utilization for each sample taken during the report period for the selected router for each 8 16 Report Descriptions ReportPack Guide buffer size There are five buffer sizes huge large big medium and small Utilization Count Buffer Utilization houston Acme s Support Office Mon Dec 20 12 00 AM Mon Dec 20 11 15 PM 138 30 92 20 Percent 46 10 o 05 00 11 00 17 00 23 00 Small Buffers Medium Buffers Big Buffers Large Buffers Huge Buffers The Buffer Counts line graph shows the number of buffers for each sam ple taken during the report period for the selected router for each buffer size There are five buffer sizes huge large big medium and small Report Descriptions Utilization Count Buffer Counts houston Acme s Support Office Mon Dec 20 12 00 AM Mon Dec 20 11 15PM 3 456 00 N w Oo e i 1 152 00 Number of Buffers Small Buffers Medium Buffers Big Buffers Large Buffers Huge Buffers 05 00 11 00 17 00 23 00 D o S e Canad U o i x J9 NOY ODSID 8 8 Cisco Router ReportPack The Buffer Faults line graph shows the number of buffer faults for each sample taken during the repor
12. Gigabytes 10 20 Nov 08 Nov 22 Dec 06 The Hourly Volume bar chart shows the volume of all Cisco routers for each hour of the report day The baseline computed for each hour over the baseline period appears as a line over the bars on the graph to compare each hour s value for that hour on that day of the week This report reveals changes over time D o S e U o i x J21NOY ODSID 8 Report Descriptions 8 3 8 Cisco Router ReportPack Hourly Daiy Hourly Volume All Cisco Routers Sun Dec 19 1999 8 Gigabytes a o to t o 00 00 06 00 12 00 18 00 D olume in Gigabytes The Hourly Grade of Service stacked bar chart shows grade of service scores for CPU Utilization Physical Memory Utilization Buffer Utilization and Buffer Errors for each sample collected on the report day Hourly Grade of Service All Cisco Routers Sun Dee 19 1999 Critical Warning Good Excellent 00 00 06 00 12 00 18 00 Di buffer Errors E Buffer Utilization E Physical Memory Utilization E CPU Utilization 8 4 Report Descriptions ReportPack Guide The Hourly CPU Utilization line graph shows the average and 95th percentile CPU Utilization of all samples collected for each hour of the report day Hourly CPU Utilization All Cisco Routers Sun Dec 19 1999 86 80 57 20 Percent 28 60 0 00 00 06 00 12 00 18 00 96th Percentile CPU Util Ave
13. Report Descriptions 12 29 _ N o lt 77 ca 3 D D O a U o 2 12 System ReportPack Top Ten Report The Top Ten report lets the network manager know at a glance which systems produce the greatest throughput and exhibit the poorest health The Top Ten report identifies the top ten systems with highest ranked metric values The top ten elements ranked by total volume and grade of service are listed by highest rank as well as highest change of rank In addition to the ranks of the elements the volume and utilization values are displayed for reference The Top Ten report comprises the following components The Systems with Highest Volume summary table ranks the selected systems in ascending sequence by volume rank for the day Systems With Highest Volume Thu Oct 21 1999 Device Element Volume Volume Rank Previous R yorktown 2 12195099049 85 1 1 eisenhower 1 4667909937 96 2 3 nimitz 4 49660733456 28 3 2 yorktown 1 1226213583 37 4 5 laffey 4 1221760234 58 5 4 nimitz 2 11944531469 83 6 6 leyte 1 1137775458 31 7 7 eisenhower 2 459146832 48 8 10 hornet 1 457219636 28 9 8 wasp 1 467 197063 48 10 g K The Systems with Greatest Change in Volume summary table ranks the selected systems in ascending sequence by change in volume rank from the 12 30 Report Descriptions ReportPack Guide previous day A large change in rank identifies a volatile resource and you should investigate the reas
14. 25 00 0 00 00 06 00 12 00 18 00 00 00 Report Descriptions 12 11 _ N W lt 77 3 D D O a U o 2 12 System ReportPack The Page Faults bar chart shows the number of page faults for the selected system over each hour of the report day Page Faults eisenhower 1 Thu Oct 21 12 00 AM 35 897 20 26 922 90 17 948 60 8 974 30 Number of Page Faults 0 Thu 00 00 Thu 09 00 Thu 18 00 The Swap Utilization line graph shows the swap utilization for the selected system Swap Utilization eisenhower 1 Thu Oct 21 12 00 AM 100 00 75 00 50 00 Percent 25 00 0 00 00 06 00 12 00 18 00 00 00 12 12 Report Descriptions ReportPack Guide The Total Swaps bar chart shows the number of swaps for the selected system over each hour of the report day Total Swaps eisenhower 1 Thu Oct 21 12 00 AM Total Swaps 0 Thu 00 00 Thu 06 00 Thu 12 00 Thu 18 00 Fri 00 00 Snapshot Report The Snapshot report shows the same information as the QuickView report for a system selected from a list at the time the report is invoked For more information about selecting a System for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Note The charts for the Snapshot report are identical to the charts for the QuickView report For chart illustrations see QuickView Report on page 12 9 The Snapshot report comp
15. 2 At the end of day 2 the baseline statistic is recomputed from all the sample data collected for days 1 and 2 and stored in the summary row for day 1 thus replacing the existing day 1 value For example if 96 samples are collected for day 2 those 96 values and the 96 values collected on the previous day day 1 x are totaled and divided by the sample count now 192 96 for day 1 plus 96 s for day 2 to arrive at value ap z Day Baseline Statistic 1 ag 3 In the same fashion the baseline statistic is recomputed at the end of day 3 from all the sample data collected for days 1 2 and 3 and stored in the summary row for day 1 again replacing the existing day 1 value At this point the value a3 for the baseline summary row represents a 3 day aggregate Day Baseline Statistic 1 a3 4 The same processing is repeated at the end of each day until a baseline value has been recalculated for each day of the rolling baseline period 42 days in Baselines 2 11 2 Overview 2 12 this example Thus at the end of 42 days a single baseline value has been computed based on data collected on the previous 42 days Day Baseline Statistic 1 a49 This day baseline value because 42 days of data is used to compute it is significantly more refined than a value where only one day s samplings are summarized to compute the baseline statistic for day 1 5 When the num
16. Figure 3 37 shows the Exception Detail table which lists a router s CPU Utilization Memory Utilization Buffer Utilization and Buffer Faults for each sample taken over the report day Exception Detail houston Acme s Support Office Sun Dec 19 12 00 AM Time Period CPU Util Mem Util Buf Util Buf Fits Sun Dec 19 12 00 AM 46 00 97 63 22 90 1 00 a Sun Dec 19 12 15 AM 40 00 97 81 22 90 0 00 m Sun Dec 19 12 30 AM 34 00 98 17 22 90 0 00 Sun Dec 19 12 45 AM 31 00 98 47 22 90 0 00 Sun Dec 19 01 00 AM 23 00 98 76 22 90 0 00 Sun Dec 19 01 15 AM 26 00 98 65 22 90 0 00 Sun Dec 19 01 30 AM 29 00 98 45 22 90 0 00 v Figure 3 37 Hot Spots Exception Detail Cisco Router 3 40 Hot Spots Report ReportPack Guide Service Level Management Report The Service Level Management Report provides an instant view of latency and availability plotted against contracted service levels Managers and customers can see at a glance if service levels have been achieved for the previous day month or other time period For example the System ReportPack is based on Empire Technology s SystemEDGE product and utilizes data collected from MIB II The following components comprise the Service Level Management report for the System ReportPack Availability table see Availability Table on page 3 42 and the following drill down component Daily Availability chart see Daily Availability Drill Down on page 3 42 Hourly Availabi
17. The Executive Summary report provides CFOs CIOs and other managers an overview of the performance of corporate interface in the critical areas of volume and utilization Each chart shows key metrics aggregated for all interfaces Key indicators of performance are shown individually and combined into a Grade of Service chart to reveal interface health at a glance Note The TRENDweb Administrator can configure a LAN or WAN Report Pack installation so that the Executive Summary reports will only dis play data for devices belonging to a specific customer In Figure 3 1 the report displays information for only the devices belonging to the Acme company For more information about this feature see Displaying Data for Specific Customer Devices in Reports on page 1 20 The Executive Summary report comprises the following components The Daily Volume bar chart shows the total volume for all interfaces over the baseline period giving the executive an instant view of increases or decreases in usage over time By default the baseline period consists of the six weeks prior to and including the report date The Hourly Volume bar chart shows the total volume for each sample taken over the report period The Hourly Baseline bar chart shows the average volume for each hour of the day over the report day You can compare total volume for each hour of the report day with the baseline for that hour for the report day Report Descripti
18. which is analyzed in related charts The utilization specified is a daily average utilization The percent collisions is calculated by dividing the total number of collisions for the day by the sum of the total number of dropped and transmitted packets Problem Summary for the Day Segments with Average Utilization gt 20 or Collisions gt 15 of Good Packets Mon Dec 13 1999 RMON Probe LF Type Speed Utilization Pct Collisions bluegrass 4_LHS_GEN ethernetecsmacd 10000000 00 0 00 3774 53 fp2000 east desktalk com 1_ALL ethernetcsmacd 10000000 00 1 73 149 87 Double click on an interface to update the content of the following components with performance data for the interface 10 2 Report Descriptions ReportPack Guide The Segment Utilization line graph shows the average and peak utilization by hour for the interface selected from the Problem Summary for a Day table Segment Utilization bluegrass ALL 3 Mon Dee 13 1999 259 80 173 20 86 60 Percent Utilization 0 00 00 06 00 12 00 18 00 00 00 Peak Segment Utilization Awg Segment Utilization D e Z N D fe V D Q The Segment Collisions line graph shows the total collisions dropped events and dropped frames by hour for the interface selected from the Problem Summary for a Day table Segment Collisions bluegrass ALL 3 Mon Dec 13 1999 0 01 a k a 0 00 00 06 00 12 00 18 00 00 00 Hourly Scale Segment Collisi
19. 1 Generate a property file For more information about generating or creating a property file see Exporting Data on page 1 21 2 After you generate or create the property file edit it to define the cust _id and customer_name values in the property table By default these are assigned a value of 2 and are consider unassigned For example the property file for the Frame Relay ReportPack might look like the following after you generate it Sanjose 1 20 2 Unassigned Customer 2 Unassigned Region atlanta 1 24 2 Unassigned Customer 2 Unassigned Region atlanta 1 42 2 Unassigned Customer 2 Unassigned Region lincoln 1 19 2 Unassigned Customer 2 Unassigned Region lincoln 1 79 2 Unassigned Customer 2 Unassigned Region orlando 1 30 2 Unassigned Customer 2 Unassigned Region orlando 1 50 2 Unassigned Customer 2 Unassigned Region 3 Supply a unique integer value for cust __id However you cannot use 2 1 or 0 Using the Property Import Feature ReportPack Guide For example suppose the SanJose and atlanta targets belonged to Thunderbolt Inc and the Lincoln and orlando targets belonged to Belle Express After you edit the property file it would look like the following D U D Q Q gt 2 3 D 1 Sanjose 1 20 1 Thunderbolt Inc 2 Unassigned Region atlanta 1 24 1 Thunderbolt Inc 2 Unassigned Region atlanta 1 42 1 Thunderbolt Inc 2 Unassigned Region lincoln 1 19 2 Belle Express 2 Unass
20. 12 00 18 00 00 00 P95 Buffer Utilization Awg Buffer Utilization The Total Number of Exceptions stacked bar graph shows the number of times a Bay router exceeded a CPU memory or buffer utilization threshold during each hour of the report day Total Number of Exceptions All Bay Routers Sat Jan 01 12 00 AM Sun Jan 02 12 00 AM 50 00 37 50 Bitemory Utilization E Buffer utilization CPU Utization 25 00 12 50 0 Sat 00 00 Sat 06 00 Sat 12 00 Sat 18 00 Sun 00 00 Report Descriptions 7 5 yoed N ius o lt D fo Po xe fo Di 7 Bay Router ReportPack Forecast Report The Forecast report lists Bay routers that are within 90 days of reaching a CPU 85 memory 80 or buffer utilization 80 threshold indicating where near term correction can prevent service degradation The calculation for the threshold is from the baseline period Those Bay routers closest to the 90 day threshold are listed first The forecast data is generated using the daily 95th percentile values for CPU memory and buffer utilization Drill down charts show CPU memory and buffer utilization grade of service discards and total bytes over the baseline period Bay routers are ranked in ascending order by CPU memory and buffer utilization DTT The Forecast report comprises the following components The Estimated Days to Threshold summary table provides the number of days until CPU memory and bu
21. 22 Creating Your Own Property File 000 1 22 Importing Property Data 0 0 aaua eaaa 1 24 Removing the ReportPack 0 000 e eee eee eee 1 25 Selecting Deferred Targets or Elements for a Snapshot Report 1 27 OVCIVIOW enne riar OREERT Cacia cas 2 1 Repor Layou 2 4s bets toh Me neta hao aiatet Steele aati Aa 2 2 Standard Reports 0 c eee eee eee 2 4 Metric Classes re danicce db kk ee Sie ate Datta oh be eel bee a Reale 2 6 Forecasting reeda hte dates a eat acl a ned A ni thane dune J 2 8 Baselines 5 c0nhcinnbaats ase baeea god e deve AEN lis 2 10 RANKING coe ca madscodaaeten ete ide dle kw pada da bes oiedalt 2 13 Frequency Distribution sssri aaaeeeaa kee n ae 2 14 Analyzing TREND ReportS sasssassnnnnnnnnnnnnnnnnnnnnnn 3 1 Executive Summary Report 0 00 cece ee 3 2 Daily Volume Chart 0 000 e eee ee 3 3 Hourly Volume and Hourly Baseline Charts 3 4 Hourly Grade of Service Chart 00 000 eee 3 5 Hourly Bandwidth Utilization Graph 0 0 2 0 eee 3 6 Forecast Report 02 cece eens 3 7 Estimated Days to Threshold DTT Summary 3 7 Grade of Service Drill Down 0 000 c eee eee 3 10 Utilization Drill Down 0 0 0 000 eee 3 11 Congestion Drill Down aaua naaa aeaa 3 12 Capacity Planning Report 0 00 0 cee eee 3 13 Overutilized WAN Interfaces Summary 2000005
22. 3 14 Daily Grade of Service Drill Down 2000 000s 3 15 ReportPack Guide Daily Utilization Drill Down 0 0 0 0 e eee eee 3 16 Daily Errors amp Discards Drill Down 00000 00 ee 3 17 TOPACA ae Mare Anais eo Meigs Mos Ae i EE a Reto aad 3 17 Busiest Bay Routers 0 0 0 cece eee 3 18 Largest Bay Router Throughput Changes 3 19 Bay Routers with the Most Percentage of Discards 3 19 Bay Routers with the Greatest Change in Percentage of Discards 3 20 QuickView and Snapshot Reports 00000 c eee eee eee 3 20 Daily Grade of Service QuickView 0000 eee eee 3 22 Snapshot Selection Table Snapshot 000 3 23 Grade of Service Snapshot 00 eee eee eee 3 23 Hourly Volume a a a a a a 3 24 Hourly Congestion assau renia teraa KEPA eae 3 24 Distribution of Bandwidth Utilization Input 3 25 Distribution of Bandwidth Utilization Output 3 26 Near Real Time QuickView and Snapshot Reports 3 27 Router Selection List 0 0 00 e eee eee eee 3 29 CPU Utilization Drill Down 0 0 00 000 eee eee 3 30 Memory Utilization Drill Down 00 0000 e eee eee 3 31 Buffer Utilization Drill Down 0 0 0000 e eee eee 3 32 Hot Spots Report x y aie a Pome teu ee te el ea es 3 32 Problem Summary for the Day nassaan 00 00 e eee eee 3 33 Hou
23. 35 Out Util 36 40 Out Util 41 46 Out Util E 46 100 Out Util The Percent Errors and Percent Discards line graph shows the percent errors and percent discards for the selected interface over the samples for each hour of the report day Report Descriptions 5 9 5 LAN and WAN ReportPacks Percent Total Percent Errors and Discards boston Northeast Sales tion Dec 20 12 00 AM 5 70 3 80 Percent 1 90 0 00 00 06 00 12 00 18 00 00 00 Awg Percent Erors Awg Percent Discards LAN only The Total Errors and Discards line graph shows the total errors and total discards for the selected interface over each hour of the report day Percent Total Total Errors and Discards boston Northeast Sales Mon Dec 20 12 00 Ari 185 708 80 139 281 60 92 854 40 Total 46 427 20 0 00 00 06 00 12 00 18 00 00 00 Total Errors Total Discards 5 10 Report Descriptions ReportPack Guide The Availability line graph shows the percent availability of the selected interface over each hour of the report day Availability is essentially computed as the ratio of the time that an interface is functional to the total time the interface is up displayed as a percentage See Daily Availability Drill Down on page 3 42 for a discussion of the factors that are used to compute availability syoequoday a gt Z o 5 a gt Z Availability boston
24. 50 Number of Frames 0 00 00 06 00 12 00 18 00 00 00 Frames with FECNS Frames with BECNS The Exception Detail table shows the number of In Utilization Out Utilization and Congestion exception conditions occurring for each sample period on the report day for the selected PVC An In Utilization or Out Utilization metric greater than 80 or Congestion metric greater than 25 is treated as an exception and is included in the chart Exception Detail orlando Accounting Server Wed Dec 22 12 00 AM Time Period Interface DLCI Congestion In Util CIR In Util Out Util CIR Out Util FECHs BECHs lfSpeed fSpeed Wed Dec 22 12 00 AM 1 50 0 00 121 24 26 87 120 63 26 71 a Wed Dec 22 12 15 AM 1 50 0 00 112 56 24 95 104 36 23 13 Wed Dec 22 12 30 AM 1 50 0 00 95 76 21 23 91 79 20 35 Wed Dec 22 12 45 AM 1 50 0 00 79 00 17 51 87 25 19 34 Wed Dec 22 01 30 AM 1 50 0 00 80 40 17 82 80 95 17 94 Wed Dec 22 01 45 AM 1 50 0 00 89 96 19 94 86 51 19 18 Wed Dec 22 02 00 AM 1 50 0 00 94 84 21 02 95 21 21 10 Wed Dec 22 02 15 AM 1 50 0 00 100 42 22 26 99 92 22 15 Ra Report Descriptions 4 17 A TI o 3 D D o o lt D o xo fe 1 4 Frame Relay ReportPack Top Ten Report The Top Ten report lets the network manager know at a glance which PVCs produce the greatest volume and the poorest health The top ten elements ranked by total volume and grade of service are listed by highest rank as well as highest change of
25. 8 15 Capacity Planning Report 02 0 e eee eee eee 8 19 Hot Spots Report cse serie a a a ees 8 23 Top Tep Repor scree ee eee es A ie ee DES eae ate 8 25 Service Level Management Report 200 eeu 8 26 Data SOUICE Ms Oraa a ea As E Ao ek ln ait bad a dd Sei dee 8 30 Basic Metric Calculations 0 000 cee ee 8 34 Grade of Service Calculation 0 00 00 cee eee 8 36 vii Contents viii RMON ReportPack 0 00 e cece eee eee eens 9 1 Understanding Remote Monitoring RMON 000 9 2 Local Montori 3 4 seckdng hed re kanen peg eo ede bag dees 9 2 Remote Monitoring 000 i en e en aE EAEE bR es 9 2 Report Descriptions 0 0 0 LaS aa AE a A a 9 3 QuickView Reports 0 0 0c cee eee eee eee 9 3 Daily RMON Ethernet 0 0 c eee eee eee 9 3 Monthly RMON Ethernet 000 eee eee eee 9 5 Daily RMON Token Ring 0 e 0c eee eee eee eee 9 7 Monthly RMON Token Ring 000ee eee eee eee 9 9 RMON2 ReportPack 00 0c eee eee 10 1 Report Descriptions 00 0 c eee 10 1 Hot Spots Report 0 a a a tee 10 2 QuickView and Snapshot Reports 000 ee eee 10 5 Daily QuickView Report 0 00 e eee eee eee 10 5 Hourly QuickView Report 00 00 eee eee eee 10 8 Snapshot Reports 00 eee eee 10 9 Top Ten Report 2 0 0 c eee eee eee es 10 9 Daily Top
26. 90 0 00 LO The Systems with the Greatest Change in Grade of Service Scores summary table ranks the selected systems in ascending sequence by change in grade of service scores from the previous day A large change in rank identifies Report Descriptions 12 31 _ N o lt 77 ca 3 D D O a U o 2 12 System ReportPack a volatile resource and you should investigate the reasons for the high change in volume Systems With Greatest Change In Grade of Service GOS Scores Thu Oct 21 1999 Device Element GOSScore Previous GOS Change Rank GOSRank Previous Rank CPU Utilization Mei GOS Score Utili leyte 4 2 45 2 43 1 3 3 51 02 5t yorktovn 4 2 25 2 26 2 4 4 34 21 5i hornet 1 1 27 1 26 3 6 6 11 66 2 nimitz 1 2 47 2 46 4 4 51 41 5t laffey 4 2 24 2 24 5 5 5 34 09 5i eisenhower 1 1 18 1 18 6 7 7 9 40 2z enterprise 4 41 17 1 18 7 8 8 9 46 a wasp 1 2 45 2 45 8 2 2 51 01 5t O Service Level Management Report 12 32 The Service Level Management report lets CIOs CFOs network managers and cus tomers know if their systems are meeting contracted service levels for availability and response times by reporting system availability across the enterprise over time and for individual Systems for the day It reports the ten systems with lowest avail ability for the day and the ten systems with the highest network response times The System Service Level Management report provides drilldowns for daily availabili
27. Accounting Server Thu Dec 23 09 00 Ai Thu Dec 23 12 15 PM 43 60 32 70 21 80 10 90 0 09 00 09 30 10 00 10 30 11 00 11 30 12 00 Sent Util IF Speed Rec Util IF Speed The PVC FECNs BECNSs line graph plots the number of frames with Forward Explicit Congestion Notification and the number of frames with Backward Explicit Congestion Notification for the selected PVC for each sample taken during the report day PY C BECNs FECNs orlando Accounting Server Thu Dee 23 09 00 Al Thu Dec 23 12 15 PM 16 00 12 00 8 00 4 00 0 09 00 09 30 10 00 10 30 11 00 11 30 12 00 BECNs FECNs Report Descriptions 4 11 yoed A TI o 3 D D ee iY lt D me fe L 4 Frame Relay ReportPack Capacity Planning Report The Capacity Planning report shows the most overutilized and underutilized PVCs based on a 90 day forecasted value Overutilized and underutilized PVCs are contrasted on this report to point out the feasibility of routing some traffic from the overutilized PVCs to the underutilized PVCs to balance the workload Drill down charts show daily data for average in and out utilization grade of service and congestion for the previous 42 days The Capacity Planning report comprises the following The Overutilized PVCs table ranks the PVCs in descending order by In Utilization and Out Utilization 70 threshold for 30 60 and 90 days Table 4 2 describes how In and Out Util
28. Northeast Sales Mon Dec 20 12 00 Aw 100 00 80 00 60 00 Percent Availability 20 00 0 00 00 06 00 12 00 18 00 00 00 Snapshot The Snapshot report comprises the following components The Snapshot Selection table lists the selected interface and its metric information Snapshot Selection Mion Dec 20 1999 Mon Dec 20 1999 Time Period Device Element Description Utilization Discards Errors Bytes Double click on an interface in the table to update the following components with performance data for the selected interface Report Descriptions 5 11 5 LAN and WAN ReportPacks 5 12 The Grade of Service stacked bar chart shows the GOS scores for the selected interfaces The chart shows weighted values for the metrics for the report day For LAN interfaces the metrics are as follows Utilization Errors and Discards For WAN interfaces the metrics are as follows Input Utilization Output Utilization Percent Errors and Percent Discards For a description of how GOS is calculated for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively LAN only The Hourly Volume bar chart shows the volume of data flow for each sample period of the report day for the selected LAN interface An overlay line graph shows the volume computed for each hour of the over the baseline period so you can compare the volume shown for the report day with the baseline LAN only The Distribution of Bandwidth Utilization pie chart
29. QuickView and Snapshot Reports The Near Real Time Quick View report provides PVC performance statistics up to the last SNMP poll The report does not rely on nightly summaries therefore it provides instant reporting on collected data Network managers and analysts can go to this report to select a PVC from a table and assess its status in the following performance areas utilization CIR and ifSpeed volume and FECNs and BECNs Alternatively the Snapshot report shows the same information however you select PVCs you want to view from a pick list when you invoke the report For more information about selecting a PVC for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The QuickView and Snapshot reports comprise the following charts The PVC Selection List table shows each PVCs and its performance statistics based on the data polled for the last 6 hours Depending on the type of statistic the value shown is either the average or total value of all the data collected over the 6 hour period PVC Selection List Select interface to see near real time reports Device IFDLCI Sent Util CIR Received FECNs BECHs SentUtil Received Util Sentkbytes Receiv Util CIR IF Speed IF Speed 78 00 portland 1 21 35 00 portland 1 30 66 69 67 40 78 00 36 00 7 39 7 47 408584 torrance 1 89 66 69 67 47 79 00 37 00 7 39 7 48 407444 torrance 1 90 46 20 42 72 77 00 36 00 5 01 4 7
30. Router 3 33 Hot Spots Hourly Grade of Service Drill Down Cisco Router 3 34 Hot Spots CPU Utilization Drill Down Cisco Router 3 36 Hot Spots Buffer Utilization Drill Down Cisco Router 3 37 Hot Spots Memory Utilization Drill Down Cisco Router 3 38 Hot Spots Percent Discards Drill Down Cisco Router 3 39 Hot Spots Buffer Faults Drill Down Cisco Router 3 40 Hot Spots Exception Detail Cisco Router 3 40 Service Level Availability Table System 3 42 Service Level Daily Availability Drill Down System 3 43 Service Level Hourly Availability Drill Down System 3 44 Service Level Response Time Drill Down System 3 45 Service Level Daily Response Time Drill Down System 3 46 Service Level Hourly Response Time Drill Down System 3 47 Tables 1 1 2 1 2 2 2 3 4 1 4 2 4 3 5 1 5 2 5 3 5 4 5 5 6 1 6 2 6 3 6 4 7 1 7 2 7 3 8 1 8 2 8 3 TREND ReportPack Dependencies 22 022000e ees 1 2 Standard Reports 000 e eee eee eee 2 4 Metric Classes 0 0 00 e eect ete tee ees 2 6 Sample Calculation for Frequency Distribution 2 14 Frame Relay Variables 0 0 00 e eee eee 4 25 Frame Relay Metrics 0 0 00 eee eee eee 4 26 Frame Relay GOS Calculations 4 28 LAN and WAN Connectivity Variables
31. SOUICE e oedd dud pe hte hee Oe eke HEP eee ead tees 4 25 Basic Metric Calculations 00 eee 4 26 Grade of Service Calculation 0 00 cee eee 4 28 LAN and WAN ReportPacks 0 eee e eee eee eee eee 5 1 Report Descriptions 0 0 00 eee ees 5 2 Executive Summary Report 0 20 5 2 Forecast Reports cos aa va As lee ea ae ee ce A 5 3 QuickView and Snapshot Reports 0 2 00 cee eae 5 6 QUICK VIOW casas ntl Gta Soe aa aa a aoe leat E Bath added A HH wae 5 6 Snapshot reeder ee ee Ee ee Wa ea eed 5 11 Near Real Time QuickView and Snapshot Reports 5 13 Near Real Time QuickView 00000 eee eee eee 5 13 Near Real Time Snapshot 20 00 eee eee eee 5 15 Capacity Planning Report 0 0 e eee eee 5 16 Hot Spots Report 00 cece eee 5 17 Top Ten Report 200 c eee eee ees 5 21 Service Level Management Report 0 00 ee 5 23 Data SOUICE oea ek ee ed eee ee ee 5 27 Basic Metric Calculations 0 0 0 e eee 5 31 Grade of Service Calculation 0 000 c eee eee 5 36 Router ReportPack 000 e cece eee 6 1 Report Descriptions 00 000 eee ees 6 2 Executive Summary Report 0 200 eee ee eee eee 6 2 Forecast Report 20 ccc eee te tee teens 6 5 ReportPack Guide QuickView and Snapshot Reports 000 eee ee eee 6 8 HOt Spots REPOM co 3424 ered Mie
32. Service Scores Wed Dec 22 1999 Wed Dec 22 1999 Critical Warning Good E Congestion E Out utilization Excellent Win utilization Bunepeyy eyuene Bununooow ujoouly Buynunosoy puejpod sajes puejpod Bunsypeyy asofues Bunayeyy OURO Sales asuenoy poddns asuenoy iamas Buyunosoy opuepo Figure 3 19 QuickView Daily Grade of Service Frame Relay Double click on a device in the chart to update the content of the associated drill down components 3 22 QuickView and Snapshot Reports ReportPack Guide Snapshot Selection Table Snapshot Figure 3 20 shows the Snapshot Selection table which lists the time period for which the PVC information was collected the PVC name the interface the PVC s Input Utilization CIR and ifSpeed Output Utilization CIR and ifSpeed and Volume Snapshot Selection Mon Dec 27 1999 Device Interface DLCI Description Congestion Input Utilization Input Utilization Output Utilizatior CIR fSpeed CIR Figure 3 20 Snapshot Snapshot Selection Table Frame Relay Double click on a device in the table to update the content of the associated drill down components with performance data for the device Grade of Service Snapshot Figure 3 21 shows the Grade of Service stacked bar chart which displays the selected PVC s GOS scores for each sample taken over the report day Grade of Service lincoln Accounting Wed Dec 22 12 00 Avi Critical
33. Tenusu aoe ca ede e a oe tana ee eae A at 10 10 Weekly Top Ten 2 0 cee eee 10 12 TREND Database ReportPack 000see eee ee eee en eee 11 1 Report Descriptions 0 0 ce ae 11 1 Database Usage Summary Report 000 eee 11 2 Data SOULE tx dese idee ed setdd desu ofa he etiucsyeda bigs 11 6 SDdsi_database_util Table 0 0 cc ee 11 7 SD42SDadsi_database_util_ forecast Table 11 8 Basic Metric Calculations 00 0 0 cece eee 11 11 System ReportPack 00 cece eee eee eee 12 1 Report Descriptions a a aaa 0 cee eae 12 1 Executive Summary Report 020 eee eee eee eee 12 2 Forecast Filesystems Report 000 eee eee eae 12 4 Forecast Memory Report 0 0c eee ee eee 12 8 QuickView Report 20 000 cece tenes 12 9 ReportPack Guide Snapshot Repor sra nese eee nE wh ea ieee REER en A 12 13 Near Real Time QuickView and Snapshot Reports 12 14 Capacity Planning Filesystems Report 55 12 18 Capacity Planning Memory Report 12 23 Hot Spots Repo siccceta haere Sats Ma attendee eRe ae x 12 25 Top ToM REDO e eted cane Ge ERE RENAE heeeiuddea Big 12 30 Service Level Management Report 20 0 055 12 32 Data SOUICE a a a E a a a a e E a A a 12 33 Basic Metric Calculations s a a auauua anaa 12 50 Grade of Service Calculation uaaa a naana 12 51 Prope
34. The Grade of Service stacked bar chart shows the GOS scores for the selected systems The chart shows weighted values for CPU memory and swap Report Descriptions 12 9 _ N w lt P 3 D O U Q 12 System ReportPack utilization and queue length for the report day See Table 12 7 for a description of the GOS calculation formulas Grade of Service Systems With Highest Grade Of Service GOS Scores Thu Oct 21 1999 Critical Warning Good E Queue Length E swap Utilization memory Utilization HE CPU Utilization Excellent asudaqua e w 2 gt 2 a Double click on a system to update the following components with performance data for the selection The Number of Processes line graph shows the number of processes for each sample period of the report day for the selected system Number of Processes eisenhower 1 Thu Oct 21 12 00 AM 446 00 334 50 223 00 111 50 Number of Processes 0 00 00 06 00 12 00 18 00 00 00 12 10 Report Descriptions ReportPack Guide The CPU Utilization line graph shows the CPU utilization for the selected system CPU Utilization eisenhower 1 Thu Oct 21 12 00 AM 400 00 76 00 j 50 00 v a 25 00 a 00 00 06 00 42 00 48 00 00 00 The Memory Utilization line graph shows the memory utilization for the selected system Memory Utilization eisenhower 1 Thu Oct 21 12 00 AM 100 00 76 00 50 00 Percent
35. Time a managed element is functional divided by the entire time period uptime for the managed element 2 of 3 Metric Classes 2 7 2 Overview Table 2 2 Metric Classes Metric Description Frequency Distribution Obtained by dividing a sample s frequency in a defined class bucket by the total number of samples For example frequency distribution can be depicted with a pie chart showing the time that an element spends in a specified utilization range over a day 3 of 3 Each metric must be interpreted in terms of the technology area to which it applies For example utilization is shown in terms of bandwidth for Frame Relay PVCs and in terms of CPU memory and buffer utilization for routers For an explanation of how the metric is applied to a specific technology see the subsequent chapter that describes that technology s ReportPack Forecasting 2 8 Forecasts are used in the Capacity Planning and Forecast reports where resource usage is forecast for 30 60 and 90 day periods and the number of days until a specified usage threshold DTT or days to threshold forecasting is reached The 30 60 and 90 day forecasts answer the following question What will the metric value for example percent utilization be for a resource in 30 60 or 90 days The days to threshold forecast answers the following question In how many days will the resource being forecast be at the thres
36. Z D UV D O QuickView Reports The QuickView reports provide daily and monthly information on individual Ethernet or Token Ring networks The daily reports help you identify potentially critical short term problems on your network The monthly reports help you evaluate utilization trends for your networks Daily RMON Ethernet The Daily RMON Ethernet report lists each Ethernet interface with its correspond ing average and peak utilization and average and peak percent collisions for the pre ceding day It comprises the following components The Daily RMON Ethernet Statistics table provides a list of all Ethernet interfaces with their daily average and peak utilization and average and peak percentage collisions The interfaces are sorted in descending order based on Report Descriptions 9 3 9 RMON ReportPack daily peak utilization Percent collisions is calculated by dividing the number of collisions by the number of good packets Daily RMON Ethernet Statistics Sorted by Peak Utilization Wed Oct 20 1999 RMOH Probe LF Avg Util Peak Util Avg Pct Peak Pct Collisions Collisions fms east desktalk com 41_ALL 1 95 23 64 0 01 0 02 tp2000 east desktalk com 1_ALL 1 90 22 52 2 23 0 74 tp2000 east desktal k com 2_ALL 1 89 22 42 0 44 0 37 sanford 4_ALL 3 04 14 05 0 02 0 10 bluegrass 3_ALL 0 39 1 13 0 00 0 00 bluegrass 110_BIl_ALL 0 39 1 13 0 00 0 00 bluegrass 110_DTE_ALL 0 21 0 80 0 00 0 00 bluegrass 21_A
37. a preceding version of the ReportPack before it completes the upgrade The upgrade installation preserves any existing data however it installs new RPTs scripts and TEEL files and it performs any required database schema updates Autopilot will show that the V4 0 ReportPack is installed and that the upgrade package was installed Note The upgrade installation is only available for pre V3 5 RMON and RMON2 ReportPacks If you have other pre V3 5 ReportPacks remove them and install the V4 0 ReportPacks available with TREND V3 6 When you upgrade from these pre V3 5 ReportPacks to V4 0 Report Packs all of your existing data is no longer supported Note Do not remove the upgrade package unless you also intend to remove the ReportPack Note The old TRENDsheet qss and TRENDgraph qgr reports are no longer supported You can manually remove the directories containing these files from DPIPE_HOME reports directory This ensures that they will not be displayed in the list of available TRENDview PRO reports Upgrading from a V3 4 x RMON or RMON2 ReportPack To upgrade from a V3 4 x RMON or RMON ReportPack l 2 Edit the install pkg file in the following directory DPIPE_HOME packages lt ReportPack gt UPGRADE_ lt ReportPack gt ap Comment out the dependency and version lines for example Installing ReportPacks ReportPack Guide Identification report_pack UPGRADE_RM
38. assigned to each metric A value of 1 corresponds to an Excellent rating Utilization Drill Down Figure 3 13 shows the Daily Utilization line graph for overutilized interfaces which allows you to review the changes and trends in utilization over the baseline period so you can immediately understand whether the overutilized interfaces represent a brief period of heavy traffic or a long standing situation of steadily growing traffic Since both In and Out Utilization in Figure 3 13 increase over time you should consider a PVC upgrade in the future Daily Utilization lincoln WVAN Link Lincoln Tulsa Tue Nov 09 1999 Mon Dec 20 1999 135 90 90 60 Teese as m Percent 45 30 0 Nov 09 Nov 23 Dec 07 P95 In Utilization Awg In Utilization P95 Out Utilization Awg Out Utilization Figure 3 13 Capacity Planning Daily Utilization Drill Down WAN Capacity Planning Report ReportPack Guide Daily Errors amp Discards Drill Down Figure 3 14 shows the Daily Errors and Discards line graph for overutilized inter faces The ratio of errors to total packets inErrors inPackets and outErrors outPack ets is calculated for each polling cycle A similar calculation is done to reveal the ratio of discards to total packets The chart shows the daily 95th percentile of these calculations as a percentage of total packets to assist you in determining whether a sudden spike has occurred or a long standing low level problem ha
39. baseline period and the 90 day forecasted P95 utilization for both in and out utilization Estimated Days to Threshold DTT Projected To Exceed Thresholds Within 90 Days All Values Based On 95th Percentile Device Interface DLCI DTT In Current P95In Projectedin DTT Out Current P95 Out Projected Out Utilization Utilization Util 90 Days Utilization Utilization Util 90 Days cL E l portland 1 30 220 00 176 03 211 02 217 00 180 04 211 97 torrance 1 88 3 00 88 34 122 18 22 00 94 61 130 31 sanjose 1 40 3 00 88 49 123 42 23 00 95 64 128 84 torrance 1 89 225 00 180 22 209 42 210 00 175 58 213 80 torrance 1 30 4 00 87 47 125 44 22 00 97 31 130 19 portland 1 21 210 00 178 13 213 26 218 00 177 84 212 01 orlando 1 50 1000 00 366 85 377 01 1000 00 369 21 384 86 orlando 1 30 61 00 57 27 93 89 63 00 58 00 92 79 Figure 3 6 Forecast Estimated Days to Threshold Table Frame Relay The managed elements are sorted with those representing the most immediate risk listed first TREND correlates multiple potential problems into one concise indication of overall risk by weighing the DTT values for all metrics For example an element with multiple metrics that are each two days from breaching a threshold is ranked as a more severe risk than an element with just one metric that is only one day from breaching a threshold When a metric is rising a DTT forecast is automatically produced If the DTT is less than 1 000 days it is recorded if the DT
40. be repeated Hourly Grade of Service Drill Down 3 34 Figure 3 31 shows the Hourly Grade of Service stacked bar chart which presents a composite grade that correlates a number of important service quality metrics appropriate to that element or resource Hourly Grade of Service houston Acme s Support Office Sun Dec 19 1999 Critical Warning M butter Faults Good E Buffer Utilization Excellent a Physical Memory Utilization W CPU Utilization Sun 00 00 Sun 06 00 Sun 12 00 Sun 18 00 Mon 00 00 Figure 3 31 Hot Spots Hourly Grade of Service Drill Down Cisco Router Hot Spots Report ReportPack Guide Health scores for each metric are assigned and weighted according to their relative importance added to other metric scores and a final grade is calculated By transforming numerous service related statistics into an easily understood stacked bar graph and presenting it over the critical troubleshooting period of the last 24 hours this graph helps you recognize changes in Cisco router health and the root cause of those changes The Grade of Service GOS metrics for Cisco routers are CPU Utilization Physical Memory Utilization Buffer Utilization and Buffer Faults A health score is assigned to each metric A value of 1 corresponds to an Excellent rating Because not all metrics may impact service quality equally each metric is multiplied by a weighting factor By assigning a grade and corresponding description for example E
41. each overutilized interface is linked to multiple drill down components immediately below the table Daily Grade of Service Drill Down Figure 3 12 shows the Daily Grade of Service stacked bar chart which presents a composite grade that correlates a number of important service quality metrics appropriate to the managed element or resource 2 Daily Grade Of Service lincoln WAN Link Lincoln Tulsa gt Tue Now 09 1999 Mon Dec 20 1999 Critical J lt ON Warning D oS Good Q Excellent 0 J NovO9 Now 13 Nov 17 Nov21 Nov25 Novw29 Dec03 DecO Dec 11 Dec15 Dec19 pw M Discards o E Errors E Out utilization W in Utilization Figure 3 12 Capacity Planning Daily Grade of Service Drill Down WAN The Health scores for each metric are assigned and weighted according to their relative importance added to other metric scores and a final grade is calculated If samples are taken every 15 minutes a daily Grade of Service GOS grade represents the average of the weighted composite scores for 96 daily samples By Capacity Planning Report 3 15 3 Analyzing TREND Reports Daily transforming numerous service related statistics into an easily understood stacked bar graph and presenting it over the baseline period the GOS chart helps you recognize changes in WAN interface health and the root cause of those changes The WAN GOS metrics are Input Utilization Output Utilization Errors and Discards A health score is
42. example IP gateways 7 end to end for example IP hosts 7 applications for example mail relays For systems including OSI protocols layers 5 and 6 may also be counted 3 of 3 Data Source 6 25 o gt D fe c D D me fe V o ie x 6 Router ReportPack Table 6 3 lists the variables polled from the MIB II ipAddrEntry table Table 6 3 Variables Polled from MIB II ipAddrEntry Table Variable OID String Base Table Column Name ipAdEntAddr 1 3 6 1 2 1 4 20 1 1 ipadentaddr009 The IP address to which this entry s addressing information pertains ipAdEntIfIndex 1 3 6 1 2 1 4 20 1 2 ipadentifindex010 The index value which uniquely identifies the interface to which this entry is applicable The interface identified by a particular value of this index is the same interface as identified by the same value of ifIndex ipAdEntNetMask 1 3 6 1 2 1 4 20 1 3 ipadentnetmask011 bits set to 0 The subnet mask associated with the IP address of this entry The value of the mask is an IP address with all the network bits set to 1 and all the host s ipAdEntBcastAddr 1 3 6 1 2 1 4 20 1 4 padentbcastaddr012 interface The value of the least significant bit in the IP broadcast address used for sending datagrams on the logical interface associated with the IP address of this entry For example when the Internet standar
43. figures that will handle the majority of spikes in network traffic A forecast that is based on simple average utilization does not adequately show when normal spikes will overload the network and cause lost productivity and lower service levels to end users Figure 3 7 shows a compressed sample baseline data set of only 3 weeks which is used to illustrate the difference between forecasts based on 95 percentile and average utilization The gsth percentile forecast clearly gives a better planning range for bandwidth that will handle more of the fluctuations in traffic demands 100 90 80 70 BO tagt 50 sath 40 30 20 10 4 0 Utilization suod y wo gt Ss a lt S gt Q J m z sj Average Utiiz ation 965th Percentile Ltiization icine Forecast Based on Average eeeeeeee Forecast Based on 95th Wie 5th Figure 3 7 Forecasts Based on 95 Percentile vs Average Utilization Forecast Report 3 9 3 Analyzing TREND Reports To assist you in analyzing long term trends each Frame Relay PVC projected to exceed exception thresholds is linked to multiple drill down components immediately below the Estimated Days to Threshold DTT table Grade of Service Drill Down Figure 3 8 shows the Grade of Service stacked bar chart which presents a composite grade that correlates a number of important service quality metrics appropriate to the managed element or resource Grade of Serv
44. following components with performance data for the selected interface The Daily Grade of Service stacked bar graph shows GOS scores for the LAN metrics Utilization Errors and Discards and WAN metrics Input Utilization Output Utilization Total Percent Errors and Total Percent Discards over the baseline period for the selected interface For more information about the GOS weighting calculations for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively The Daily Utilization line graph shows the average and 95th percentile utilization over each day of the baseline period for the selected interface The Daily Errors and Discards line graph shows the average and 95th percentile errors and discards for the selected interface for each day of the baseline period Values are shown as percentages Hot Spots Report The Hot Spots report provides the network manager and other members of the net work management staff a list of interfaces that are exceeding threshold conditions The Hot Spots report comprises the following components The Problem Summary for the Day summary table displays the top ten interfaces in descending order by total number of exceptions An exception occurs when an interface exceeds a metric threshold in any sample taken during the report day Report Descriptions 5 17 syoequodey o m gt z D a gt zZ 5 LAN and WAN ReportPacks Problem Summary for the Day LAN Inter
45. for which the PVC information was collected the PVC name the interface the PVC s Input Utilization CIR and ifSpeed Output Utilization CIR and ifSpeed and Volume Double click on a device in the table to update the content of the following drill down components with performance data for the device Snapshot The Grade of Service stacked bar chart shows the selected PVC s GOS scores for each sample taken over the report day The Hourly Volume area graph shows volume for the selected PVC for each sample taken over the report day The Hourly Congestion line graph plots the number of frames with Forward Explicit Congestion Notification and the number of frames with Backward Explicit Congestion Notification for the selected PVC for each sample taken during the report day The Distribution of Bandwidth Utilization Input pie chart shows the percentage of samples taken during the day in which the In Utilization metric calculated for the samples fall in a defined range For more information about frequency distribution see Frequency Distribution on page 2 14 The Distribution of Bandwidth Utilization Output pie chart shows the percentage of samples taken during the day in which the Out Utilization metric calculated for the samples fall in a defined range For more Report Descriptions ReportPack Guide information about frequency distribution see Frequency Distribution on page 2 14 Near Real Time
46. from the threshold If the threshold is too far away the forecast is meaningless If the slope is negative decreasing sloping downward the result returned for DTT is a null value displayed as Not Applicable in reports Likewise if the threshold is more than 1 000 days away the result returned is null and displayed as Not Applicable For formulas that TRENDsum uses to compute all forecasts see the section on the TRENDsum utility in Appendix A of the TREND User s Guide Baselines 2 10 A baseline provides a basis for comparing recent performance with past performance TREND ReportPacks use a rolling baseline period of 42 days 6 weeks to compute baselines In a rolling baseline the baseline table contains n rows each row represents the same statistic or statistics computed for the immediately preceding number of days in the baseline period For example if you use a 42 day baseline period the rolling daily baseline is computed as follows 1 At the end of day 1 the baseline statistic is computed from all the sample data collected for day 1 and stored in a summary row for day 1 For example assume you are computing the average value of an element property to use as a baseline If 96 samples are collected for day 1 one sample every 15 minutes Baselines ReportPack Guide the values for all samples are totaled and divided by 96 to compute the average for the day 1 Day Baseline Statistic 1 ay
47. gMiss 1 3 6 1 4 1 9 2 1 35 Contains the number of big buffer misses Changed from INTEGER to Counter by DeskTalk Systems Inc bufferLgSize 1 3 6 1 4 1 9 2 1 38 Contains the size of large buffers bufferLgTotal 1 3 6 1 4 1 9 2 1 39 Contains the total number of large buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc 3 of 5 Data Source ReportPack Guide Table 8 1 Cisco Router Variables Variable OID String bufferLgFree 1 3 6 1 4 1 9 2 1 40 Contains the number of free large buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferLgHit 1 3 6 1 4 1 9 2 1 42 Contains the number of large buffer hits Changed from INTEGER to Counter by DeskTalk Systems Inc bufferLgMiss 1 3 6 1 4 1 9 2 1 43 Contains the number of large buffer misses Changed from INTEGER to Counter by DeskTalk Systems Inc bufferFail 1 3 6 1 4 1 9 2 1 46 Count of the number of buffer allocation failures Changed from INTEGER to Counter by DeskTalk Systems Inc bufferNoMem 1 3 6 1 4 1 9 2 1 47 Count of the number of buffer create failures due to no free memory Changed from INTEGER to Counter by DeskTalk Systems Inc bufferHgSize 1 3 6 1 4 1 9 2 1 62 Contains the size of huge buffers bufferHgTotal 1 3 6 1 4 1 9 2 1 63 Contains the total number of huge buffers Changed from INTEGER to Gauge by Des
48. ifindex value of the interface and the DLCI value of the PVC concatenated and separated by a period for example 3 234 port_speed The port speed in bits second float The ReportPack requires this value Frame Relay ReportPack A 3 A Property File Formats Table A 2 Frame Relay ReportPack Property File Formats Column Name Function Value Type pvc_cir The PVC s Committed Informa float tion Rate in bits second The ReportPack requires this value cust_id Unique integer value for each integer customer customer_name Customer name associated with varchar 128 cust_id region_id Unique integer value for each integer region region_name Region name associated with varchar 128 region_id value dsi_descr A description of the PVC varchar 128 Router ReportPack Table A 3 lists the property file formats for the Router ReportPack Table A 3 Router ReportPack Property File Formats Column Name Function Value Type dsi_target_name Device name whichis the name varchar 64 you use to poll the device dsi_table_key The device index which is 0 varchar 128 A 4 Router ReportPack ReportPack Guide Table A 3 Router ReportPack Property File Formats Column Name Function Value Type cust_id Unique integer value for each integer customer customer_name Customer name associated with varchar 128 cu
49. information about selecting a Bay router for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The QuickView and Snapshot reports comprise the following components The Daily Grade of Service stacked bar graph shows the GOS scores for the selected Bay routers The chart shows weighted values for CPU memory and buffer utilization for the report day See Table 7 3 for a description of the GOS calculation formulas Daily Grade Of Service GOS by Router Sun Dec 19 1999 Sun Dec 19 1999 Critical Warning E CPU Utilization E Buffer utilization Good Bhtemory Utilization Excellent yoed N w iy lt D e i 2 od Double click on a Bay router to update the following components with performance data for the selected interface Report Descriptions 7 9 7 Bay Router ReportPack The Hourly Volume bar graph shows the volume of data for each sample period of the report day for the selected Bay router Hourly Yolume sanjose Sun Dec 19 1999 4 308 50 3 846 80 2 885 10 Gigabytes 1 923 40 961 70 0 Sun 00 00 Sun 09 00 Sun 18 00 The CPU Utilization Frequency Distribution pie chart shows the percentage of CPU utilization values computed from each sample taken during the report day that fall within the defined utilization buckets for the selected Bay router CPU Utilization Frequency Distribution sanjose 1 Su
50. must remain constant at least from one reinitialization of the entity s network management system to the next reinitialization o gt D fe c s D me e U o P x ifDescr 1 3 6 1 2 1 2 2 1 2 A textual string containing information about the interface This string should include the name of the manufacturer the product name and the version of the hardware interface ifType 1 3 6 1 2 1 2 2 1 3 The type of interface distinguished according to the physical link protocol s immediately below the network layer in the protocol stack ifSpeed 1 3 6 1 2 1 2 2 1 5 An estimate of the interface s current bandwidth in bits per second For interfaces which do not vary in bandwidth or for those where no accurate estimation can be made this object should contain the nominal bandwidth ifPhysAddress 1 3 6 1 2 1 2 2 1 6 1 of 3 Data Source 6 21 6 Router ReportPack Table 6 1 Variables Polled from MIB II ifEntry Table Variable OID String The interface s address at the protocol layer immediately below the network layer in the protocol stack For interfaces which do not have such an address e g a serial line this object should contain an octet string of zero length ifAdminStatus 1 3 6 1 2 1 2 2 1 7 The desired state of the interface The testing 3 state indicates that no operational packets can be passed ifO
51. o 2 oo zo o o p e 11 TREND Database ReportPack Table 11 2 TREND Database Baseline and Forecast Variables Variable Description F60AVGpct_dbused The 60 day projection for the average percentage of database use F60MAXkb_logused The 60 day projection for the maximum number of kilobytes the transaction log will use F60MA Xpct_logused The 60 day projection for the maximum percentage of transaction log use F9OAVGkb_dbused The 90 day projection for the average number of kilobytes the database will use F9OAVGpct_dbused The 90 day projection for the average percentage of database use FO9OMAXkb_logused The 90 day projection for the maximum number of kilobytes the transaction log will use F9OMA Xpct_logused The 90 day projection for the maximum percentage of transaction log use 3 of 3 11 10 Data Source ReportPack Guide Basic Metric Calculations The performance metrics used in TREND Database reports are based on the elements from the dsi_dbstats table listed in Table 11 3 The table shows equivalent calculations for some of the elements Table 11 3 TREND Database Metrics Metric Computed As How Computed Database prused The percentage of database usage Utilization from dsi_dbstats which is equivalent to the ratio of total bytes used to the size of the database multiplied by 100 Transaction prlogused The percentage of transaction log Lo
52. offenders ranked by total number of exceptions are presented in a summary table to assist the network management team in analyzing the problems Each exception is automatically hot linked to multiple drill down charts including grade of service CPU utilization memory utilization swap capacity and number of swaps An exception detail table reveals the times and offending levels for each exception The Hot Spots report comprises the following components The Problem Summary for the Day summary table displays the top ten systems in descending order by total number of exceptions CPU memory total number of swaps paging page fault exceptions for the report day An exception occurs when an interface exceeds its defined CPU utilization 50 memory utilization 80 total number of swaps gt 10 total number of pages Report Descriptions 12 25 _ N w lt P 3 D O U D Q 12 System ReportPack gt 30 000 or total number of page faults gt 100 000 threshold in any sample taken during the report day Problem Summary for the Day System With Most Exceptions For The Day Thu Oct 21 1999 Device Interface CPU Util Memory Util Paging Page Fault Swap Exceptions Total Exceptions Exceptions Exceptions Exceptions Exceptions 2 ee es ee a wasp 1 54 00 15 00 75 00 40 00 12 00 196 00 nimitz 14 55 00 17 00 74 00 38 00 10 00 194 00 laffey 1 14 00 22 00 67 00 0 00 48 00 152 00 yorktouun 1 16 00 18 00
53. on page 2 10 Ranking on page 2 13 and Frequency Distribution on page 2 14 2 1 N ie lt D i 2 Overview Report Layout All TREND reports have a uniform layout that makes the report easy to understand Figure 2 1 shows a sample report This report gives network managers and NOC team members a listing of the oE SKTALRK Heading gt Cisco Router ciso ruters that have exceeded threshold conditions Offending routers are T REND listed in decreasing order with highest number of exceptions listed first Hot Spots Summar SE cal Problem Summary for the Day y Cisco Routers With Highest Exception Counts Sun Dec 19 1999 Router CPU Utilization Memory Utilization Buffer Utilization Buffer Failure Total Exceptions Exceptions Exceptions Exceptions Exceptions cotati 0 00 45 00 96 00 0 00 141 00 tulsa 7 00 0 00 0 00 0 00 7 00 boston 7 00 0 00 0 00 0 00 7 00 Hourly Grade of Service houston Acme s Support Office Sun Dee 19 1999 Drill Down citoi Analysis Warning M butter Faults Good E Butfer utilization Excellent E Physical Memory Utilization W CPU Utilization Sun 00 00 Sun 06 00 Sun 12 00 Sun 18 00 Mon 00 00 Figure 2 1 TREND Report Layout 2 2 Report Layout ReportPack Guide The report components are as follows although not all reports contain drill down analysis Heading The heading identifies the report and gives a high level description of the report contents Figure
54. operational In the absence of a Data Link Connection Management Interface virtual circuit entries rows may be created by setting virtual circuit state to active or deleted by changing Circuit state to invalid Whether or not the row actually disappears is left to the implementation so this object may actually read as invalid for some arbitrary length of time It is also legal to set the state of a virtual circuit to inactive to temporarily disable a given circuit frCircuitReceivedFECNS 1 3 6 1 2 1 10 32 2 1 4 Number of frames received from the network indicating forward congestion since the virtual circuit was created frCircuitReceivedBECNS 1 3 6 1 2 1 10 32 2 1 5 Number of frames received from the network indicating backward congestion since the virtual circuit was created frCircuitSentFrames 1 3 6 1 2 1 10 32 2 1 6 The number of frames sent from this virtual circuit since it was created frCircuitSentOctets 1 3 6 1 2 1 10 32 2 1 7 The number of octets sent from this virtual circuit since it was created 1 of 2 yoed Data Source 4 25 A TI o 3 D D o o lt D o fe T 4 Frame Relay ReportPack Table 4 1 Frame Relay Variables RFC 1315 Variable OID String frCircuitReceivedFrames 1 3 6 1 2 1 10 32 2 1 8 Number of frames received over this virtual circuit since it was created frCircuitReceivedOctets 1 3 6 1 2 1 10
55. other managers an overview LAN Connectivity of the performance of corporate LANs Each chart shows key ii P R E N D EF r metrics aggregated for all LAN interfaces Key indicators of Executive Summary performance are shown individually and combined into a Grade of Service chart to reveal LAN interface health at a glance Customer Name Acme os gt 72 D N D o gt Q Daily Volume G All LAN Interfaces m Tue Nov 09 1999 Mon Dec 20 1999 o 45 20 33 90 Gigabytes N N 11 30 0 Nov09 Now13 Now17 Now24 Nov25 Now29 Dec03 DecOF Dec11 Dec15 Dec19 Figure 3 1 Executive Summary Daily Volume Chart LAN Executive Summary Report 3 3 3 Analyzing TREND Reports It also puts the most recent volume information in context with the long term historical trend that is shown over the baseline period Note The TRENDweb Administrator can configure a ReportPack installa tion so that the Executive Summary reports will only display data for devices belonging to a specific customer In Figure 3 1 the report dis plays information for only the devices belonging to the Acme company For more information about this feature see Displaying Data for Spe cific Customer Devices in Reports on page 1 20 Hourly Volume and Hourly Baseline Charts Figure 3 2 shows the Hourly Volume bar chart which details the total volume of traffic for each hour yesterday and focuses the executive s a
56. repeated here for the convenience of management stations which cannot simultaneously poll groups of objects in different branches of the MIB swapCapacity 1 3 6 1 4 1 546 1 1 7 8 30 The percentage of the system s total swap in use The value of this object can be computed by dividing swapInUse 13 by totalSwapSpace 29 and multiplying by 100 It is included here for convenient polling monitoring and history sampling This object is an aggregate value over all the swap partitions and areas memCapacity 1 3 6 1 4 1 546 1 1 7 8 31 The percentage of the system s memory in use The value of this object can be approximated by dividing activeMem 10 by memory system 3 and converting to a percentage Because virtual memory allows systems to operate with a higher memCapacity value this value alone should not be used to judge the amount of the system load Managers should also monitor swapCapacity N memInUseCapacity 1 3 6 1 4 1 546 1 1 7 8 32 cH m 8 of 9 g D xe e Data Source 12 41 ae S 12 System ReportPack Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The percentage of the system s memory in use The value of this object can be approximated by dividing memInUse 9 by memory system 3 and converting to a percentage Because virtual memory allows systems to operate with a higher memCapacity value this value alone shoul
57. requested be transmitted to a subnetwork unicast address including those that were discarded or not sent ifOutNUcastPkts 1 3 6 1 2 1 2 2 1 18 The total number of packets that higher level protocols requested be transmitted to a non unicast 1 e a subnetwork broadcast or subnetwork multicast address including those that were discarded or not sent ifOutDiscards 1 3 6 1 2 1 2 2 1 19 The number of outbound packets which were chosen to be discarded even though no errors had been detected to prevent their being transmitted One possible reason for discarding such a packet could be to free up buffer space ifOutErrors 1 3 6 1 2 1 2 2 1 20 The number of outbound packets that could not be transmitted because of errors 4 of 5 5 30 Data Source ReportPack Guide a Dr 2 p 2 Table 5 1 LAN and WAN Connectivity Variables 3 2 Variable OID String 5 Z ifOutQLen 1 3 6 1 2 1 2 2 1 21 The length of the output packet queue in packets ifSpecific 1 3 6 1 2 1 2 2 1 22 A reference to MIB definitions specific to the particular media being used to realize the interface For example if the interface is realized by an ethernet then the value of this object refers to a document defining objects specific to ethernet If this information is not present its value should be set to the OBJECT IDENTIFIER 00 which is a syntactically valid object identifier and any conform
58. shows the percentage utilization values that fall within the defined utilization buckets for the selected LAN interface for the report day The following table lists the buckets for input utilization WAN only The Distribution of Input Bandwidth Utilization and Dis tribution of Output Bandwidth Utilization pie charts show the percent age input utilization values and output utilization values respectively that fall within the defined utilization buckets for the selected WAN interface for the report day The following table lists the buckets for input and out put utilization LAN only The Percent Errors and Percent Discards line graph shows the percent errors and percent discards for the selected interface over the samples for each hour of the report day The Total Errors and Discards line graph shows the total errors ifouter rors ifinerrors inunknownprotos and total discards ifindiscards ifoutdiscards for the selected interface over each hour of the report day The Availability line graph shows the percent availability of the selected interface over each hour of the report day Availability is essentially computed as the ratio of the time that an interface is functional to the total time the interface is up displayed as a percentage See Daily Availability Drill Down on page 3 42 for a discussion of the factors that are used to compute availability Report Descriptions ReportPack Guide Near Real Time QuickView
59. started Changed from INTEGER to Counter by DSI wfResourceTotalMemory Used 1 3 6 1 4 1 18 3 3 2 5 7 1 5 The current number of bytes of memory used by all the entities on the slot wfResourceTotalMemoryFree 1 3 6 1 4 1 18 3 3 2 5 7 1 6 The current number of bytes of free memory on the slot wfResourceTotalMemoryMax 1 3 6 1 4 1 18 3 3 2 5 7 1 7 The maximum amount of Memory available in bytes on the slot yoed wtResourceTotalBuffersUsed 1 3 6 1 4 1 18 3 3 2 5 7 1 8 The current number of buffers used by all the entities on the slot N ius o lt D fo Ps xe fo Di wfResourceTotalBuffersFree 1 3 6 1 4 1 18 3 3 2 5 7 1 9 The current number of free buffers on the slot wfTotalBufersMax 1 3 6 1 4 1 18 3 3 2 5 7 1 10 The maximum number of buffers on the slot 2 of 2 Data Source 7 27 7 Bay Router ReportPack Basic Metric Calculations The performance metrics used in Bay router reports are based on the calculations listed in Table 7 2 Table 7 2 Bay Router Metrics Metric Computed As How Computed CPU rcetotalcpuused 100 The ratio of the amount of CPU time Utilization urcetotalcpumax 100 used collected as centiseconds and 100 multiplied by 100 to yield seconds to the maximum amount of CPU time available collected in centiseconds and multiplied by 100 to yield seconds is multiplied by 100 to yield a percent
60. table shows all the routers used in the enterprise the number and types of interfaces LAN WAN or other the number of routers managed and the volume of the routers Router Inventory Thu Oct 21 1999 Vendor WAH Interfaces LANH Interfaces Other Total Routers Managed Volume in Interfaces Interfaces Gigabytes All Devices 5 6 0 11 4 79 21 The Daily Volume bar chart displays the total router volume for all routers e2 over the baseline period this gives the executive an instant view of increases D or decreases in router usage over time By default the baseline period consists of the six weeks prior to and including the report date wy Ay Daily Yolume ks All Routers Q Fri Sep 10 1999 Thu Oct 21 1999 5 o Q 94 20 x 78 50 62 80 a 47 40 2 31 40 15 70 0 Sep 10 Sep 24 Oct Os Report Descriptions 6 3 6 Router ReportPack The Hourly Volume bar chart displays the total router volume over the day specified and compares each hour s value to the average value for that hour over the baseline period This chart also shows changes over time Volume Basetine Hourly Volume 5 50 4 40 Gigabytes wo io o Lee iv 1 10 0 00 00 06 00 12 00 18 00 6 4 Report Descriptions ReportPack Guide The Baseline Volume line graph shows the baseline volume of the devices over the baseline period Yolume BaseLine Baseline Volume 5 50 4 40 o gt o F 3 30 E E 5
61. the PVCs sorted by change in rank in ascending order based on the report day s volume change PVCs With Greatest Change In Volume Wed Dec 22 1999 Device Interface DLCI Description Volume Previous Volume Volume Volume Rank Pre Change Rank portland 1 30 Sales 166121302 16 166689312 71 1 4 orlando 1 50 Accounting Sewer 687821902 63 88363836 02 2 7 portland 1 24 Accounting 166497769 61 166096161 74 3 3 torance 1 90 Marketing 87768136 22 87419287 99 4 6 torance 1 89 Support 166539485 82 1663907 17 66 5 2 torrance 1 88 Sales 87585468 15 87684576 72 6 8 sanjose 1 40 Marketing 87698904 20 87625234 06 7 7 orlando 1 30 Accounting 60177092 68 60137201 84 8 9 9 6 Sanjose 1 20 Support 87762855 55 87744522 20 An advantage to using this table is that it may highlight which newly installed devices and technologies are approaching a logical threshold may be exceeding a threshold or may be impeding optimum utilization The PVCs with Highest Grade of Service GOS Scores table shows PVCs sorted by rank in ascending order based on the report day s GOS Score Table 4 3 describes how Grade of Service is calculated PVCs With Highest Grade Of Service GOS Scores Wed Dec 22 1999 Device Interface DLCI Description GOS Score GOSRank Previous Rank Utilization Congestion v lincoln 1 19 Accounting 2 16 4 4 52 77 4 19 330 atlanta 1 42 Marketing 2 16 2 4 52 75 4 19 330 orlando 1 50 Accounting Server 1 81 3 2 469 77 0 00 5874 portland 1 30 Sales
62. the selected router Bay Router Daily Memory Utilization sanjose Desktalk Support 1 Tue Dec 21 1999 Mion Jan 31 2000 100 00 75 00 50 00 Percent 25 00 0 Dec 21 Jan 04 Jan 18 Avg hemory Utilization P95 Memory Utilization 7 16 Report Descriptions ReportPack Guide The Underutilized Bay Routers summary table shows the Bay routers that projected to be the least utilized within the next 90 days ranked in ascending sequence by the 90 day projected CPU memory or buffer utilization value All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Underutilized means that the router is furthest from reaching the utilization thresholds 10 defined Underutilized Bay Routers Projected To Be Least Utilized In 90 Days Thresholds CPU Util 10 Memory Util 10 Buffer Util 10 Router Slot 30Day CPU 60 Day CPU 90 Day CPU 30DayMemory 60DayMemory 90DayMemory 30 Util Forecast Util Forecast Util Forecast Util Forecast Util Forecast Util Forecast Util sanluis 41 7 80 7 80 7 80 100 00 100 00 100 00 sanluis 2 8 58 8 58 8 58 32 07 32 11 32 14 UO Double click on a Bay router in this table to update the following components with performance data for the selected router The Bay Router Daily Grade of Service stacked bar graph shows GOS scores for CPU memory and buffer utilization over the baseline period for the selected router See Table 7 3 fo
63. the top end of currently acces sible resources additional investment may be required to satisfy end user demands Forecast Report The Forecast Report delivers a view into the future so that proactive corrections can be made to prevent service problems The Forecast report is also a valuable resource to plan network expansion and justify future capital expenditures For example the following components comprise the Forecast report for Frame Relay Estimated Days to Threshold DTT summary table see Estimated Days to Threshold DTT Summary on page 3 7 and the following drill down components Grade of Service chart see Grade of Service Drill Down on page 3 10 Utilization chart see Utilization Drill Down on page 3 11 Congestion chart see Congestion Drill Down on page 3 12 The following sections describe these charts in more detail For detailed information about the Frame Relay ReportPack see Frame Relay ReportPack on page 4 1 swioday i gt lt Ss Ko J m z s Estimated Days to Threshold DTT Summary Figure 3 6 shows the Estimated Days to Threshold DTT summary table which provides a ranked list of all elements that have exceeded or are forecasted to be within 90 days of reaching a threshold exception condition The DTT is shown along Forecast Report 3 7 3 Analyzing TREND Reports 3 8 with the average of the daily P95 utilization over the
64. the3DPIPE_HOME tmp trend 1og file for details During Key Table Upgrade stage the upgrade program tries to recreate indexes on the key tables This process requires a large amount of t emp_db and database log If the upgrade fails during index recreation because there is not enough space you must recreate the index manually after additional space is allocated To recreate index run the following command isql Udsi_dpipe Pdsi_dpipe 1 gt create unique index K2 on ghost_key_tablename 2 gt dsi_control_table_keyid dsi_table_key 3 gt go 1 gt quit Demo Installation A demo installation installs a ReportPack that has example reports Each ReportPack provides a demo package for example Syst em_Demo ap which you can install When you install the demo ReportPack example report files are installed in the following directory DPIPE_HOME reports Demo lt ReportPack gt subdirectory In this directory the report files are visible to all TRENDweb servers and all TRENDview PRO clients Installing ReportPacks 1 9 1 ReportPack Administration Note You can view the demo reports sdt files without a TREND system using TRENDview Pro If you prefer to do this do not use Autopilot to install the demo use TRENDview PRO to open the sdt files For more information about TRENDview PRO see the Guide to Viewing TREND Reports Running Autopilot The installation of a ReportPack using Autopilot C
65. though no errors had been detected to prevent their being transmitted One g possible reason for discarding such a packet could be to free up buffer V space D ifOutErrors 1 3 6 1 2 1 2 2 1 20 The number of outbound packets that could not be transmitted because of errors 3 of 3 Table 6 2 lists the variables polled from the MIB II system table Table 6 2 Variables Polled from MIB II System Table Variable OID String sysDescr 1 3 6 1 2 1 1 1 A textual description of the entity This value should include the full name and version identification of the system s hardware type software operating system and networking software It is mandatory that this only contain printable ASCII characters 1 of 3 Data Source 6 23 6 Router ReportPack 6 24 Table 6 2 Variables Polled from MIB II System Table Variable OID String sysObjectld 1 3 6 1 2 1 1 2 The vendor s authoritative identification of the network management subsystem contained in the entity This value is allocated within the entity This value is allocated within the SMI enterprises subtree 1 3 6 1 4 1 and provides an easy and unambiguous means for determining what kind of box is being managed For example if vendor Flintstones Inc was assigned the subtree 1 3 6 1 4 1 4242 1 3 6 1 4 1 4242 1 1 to its Fred Route it could assign the identifier r sysUptime 1 3 6 1 2 1 1 3 The tim
66. to unmount this device and remove its entry from the mounted file system table Writing any other value to this object will cause no action to be taken by the agent devCapacity 1 3 6 1 4 1 546 1 1 1 7 1 14 7 3 3 of 4 T wv xe e Data Source 12 49 D pa 12 System ReportPack Table 12 5 System Variables Polled from Empire devTableEntry Table Variable OID String The percentage of the device s total capacity in use 4 of 4 Basic Metric Calculations The performance metrics used in the System reports are based on the calculations listed in Table 12 6 Table 12 6 System Metrics Metric Computed As Memory Utilization activemem meminuse 100 Total Swaps numswapins numswapouts Volume inoctets outoctets 12 50 Basic Metric Calculations ReportPack Guide Grade of Service Calculation Table 12 7 lists how the metrics are weighted in the System Grade of Service GOS charts Table 12 7 System GOS Calculations CPU Memory Swap Queue GOS Calculation Utilization Utilization Utilization Length GOS weight factor 25 25 25 25 GOS score Excellent 0 24 0 24 0 24 0 1 GOS score 2 Good 25 49 25 49 25 49 2 lt 3 GOS score 4 Warning 50 74 50 74 50 74 3 4 GOS score 5 Critical 75 100 75 100 75 100 5 or more For more information about the Grade of Service metric see Tab
67. virtual memory page that is not currently in physical memory resulting in a hardware page fault A high rage of page faults may indicate an overloaded system or one with insufficient memory loadAverage Min 1 3 6 1 4 1 546 1 1 7 8 26 The load average over the last minute The load average represents the average number of jobs in the run queue over the specified time range The value reported is the load average multiplied by 100 For example the value 50 represents a load average of 0 50 loadAverage5 Min 1 3 6 1 4 1 546 1 1 7 8 27 The load average over the last 5 minutes The load average represents the average number of jobs in the run queue over the specified time range The value reported is the load average multiplied by 100 For example the value 50 represents a load average of 0 50 loadAverage15Min 1 3 6 1 4 1 546 1 1 7 8 28 7 of 9 12 40 Data Source ReportPack Guide Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The load average over the last 15 minutes The load average represents the average number of jobs in the run queue over the specified time range The value reported is the load average multiplied by 100 For example the value 50 represents a load average of 0 50 totalSwapS pace 1 3 6 1 4 1 546 1 1 7 8 29 Total system swap space in KBytes This object is identical to the kernelConfig totalSwap 12 object and is
68. 0 01 00 AM 32 34 4 06 5 38 0 10 113 57 Mon Dec 20 01 15 AM 30 05 3 86 5 11 0 10 115 98 Mon Dec 20 02 15 PM 28 94 3 76 5 08 0 10 116 08 Mon Dec 20 02 30 PM 31 07 3 87 5 24 0 10 115 22 5 20 Report Descriptions ReportPack Guide Top Ten Report syoequoday o gt Z 2 gt a The Top Ten report lets the network manager know at a glance which interfaces produce the greatest volume and the poorest health The top ten elements ranked by total volume and grade of service are listed by highest rank as well as highest change of rank Volume and health change charts alert you to the interface s greatest changes in volume suggesting potential capacity overloading if increasing trends continue The volume change report also may highlight the newly installed devices and technologies that are approaching a utilization threshold or are underutilized Drastic increases in GOS scores suggest additional interfaces to be watched while decreases document an effective remedy to a previously negative condition For more information about the GOS weighting calculations for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively The Top Ten report comprises the following components The LAN Interfaces with Highest Volume table shows up to ten interfaces with the highest volume for the report day LAN Interfaces With Highest Volume Mon Dec 20 1999 Device Interface Description Volume Volume Rank Previous Rank Utilizati
69. 0 Marketing 0 00 21 00 33 00 54 00 54 00 LO Double click on a device to update the content of the following charts with performance data for the device The Grade of Service stacked bar chart shows the Grade of Service GOS ratings for In Utilization Out Utilization and Congestion for the Report Descriptions 4 15 yoed A TI o 3 D D o o lt D me fe 1 4 Frame Relay ReportPack selected PVC for each sample taken on the report day for the selected PVC Table 4 3 describes how Grade of Service is calculated Grade of Service GOS orlando Accounting Server Wed Dec 22 12 00 At Critical Warning Good E Congestion E Out Utilization Win utilization Excellent Wed 00 00 Wed 06 00 Wed 12 00 Wed 18 00 Thu 00 00 The Utilization line graph shows In Utilization and Out Utilization computed using both CIR and ifSpeed for the selected PVC for each sample period on the report day Utilization orlando Accounting Server Wed Dec 22 12 00 AM 412 20 274 80 Percent 137 40 0 00 00 06 00 12 00 18 00 00 00 Input Utilization CIR Output Utilization CIR Input Utilization Ifspeed Output Utilization Ifspeed Report Descriptions ReportPack Guide The Congestion line graph shows congestion for the selected PVC for each sample taken on the report day Congestion orlando Accounting Server Wed Dec 22 12 00 AM N w S E amp a Oo o o Oo 11
70. 0 Mon 18 00 Mon21 00 Tue 00 00 E Erors M Discards H utilization The Utilization line graph shows the average percentage of bandwidth utilization LAN or the input and output utilization WAN for each hour of the report day for the specified interface Utilization sanjose Sales Mon Dec 20 12 00 AM 0 00 00 06 00 12 00 18 00 00 00 Avg Utilization The Percent Errors and Discards line graph shows the average percentage of errors and discards LAN or in errors out errors in Report Descriptions 5 19 5 LAN and WAN ReportPacks discards and out discards WAN for the specified interface over the hours of the report day Percent Errors and Discards sanjose Sales Mon Dec 20 12 00 Awl 6 90 4 60 Percent 2 30 0 00 00 06 00 12 00 18 00 00 00 Awg Percent Erors Awg Percent Discards The Exception Detail table shows the time the sample was taken for each exception and the values for the offending conditions For more information about how the metrics are calculated for the LAN and WAN interfaces see Table 5 2 and Table 5 3 respectively Exception Detail sanjose Sales Mon Dec 20 12 00 AM Time Period Utilization Percent Percent Errors Pct Broadcast Avg Packet Size Discards Packets Mon Dec 20 12 00 AM 30 57 4 00 5 29 0 10 112 84 Mon Dec 20 12 15 AM 32 66 4 28 5 54 0 14 121 60 Mon Dec 20 12 30 AM 32 34 3 98 5 94 0 09 107 29 Mon Dec 20 12 45 AM 35 92 4 27 6 04 0 10 116 84 Mon Dec 2
71. 00 9 33 9 72 2 53 1 85 hiktgs tulsa 2 10000000 6 63 6 73 1 98 0 00 Tulsa Research Cent houston 2 10000000 6 44 6 46 2 01 0 00 Accounting cotati 2 10000000 1 92 1 95 1 19 0 10 Cotati Sales 7 Double click on an interface in the chart update the following components with performance data for the selected interface Report Descriptions 5 13 5 LAN and WAN ReportPacks The Interface Average Traffic Rate line graph shows the average traffic rate for each sample taken during the report period for the selected LAN interface Interface Average Traffic Rate boston Northeast Sales Mon Dec 20 01 00 Phi Tue Dec 21 01 30 PM 5 181 50 4 145 20 3 108 90 2 072 60 Traffic Rate kbps 1 036 30 0 Mon 21 00 Tue 06 00 Average Data Rate The Interface Utilization line graph shows the average utilization for the percentage utilization for each sample taken during the report period Interface Utilization boston Northeast Sales Mon Dec 20 01 00 Phi Tue Dec 21 01 30 PM 42 00 w in 5 21 00 Percent Utilization 10 50 Mon 21 00 Tue 06 00 Average Utilization 5 14 Report Descriptions ReportPack Guide The Errors and Discards line graph shows the total errors and total discards for the selected interface for each sample taken during the report period syoequoday o f gt Z o 5 a gt Z Errors and Discards boston Northeast Sales Mon Dec 20 01 00 PM Tue
72. 1 4 1 546 1 1 7 8 10 The total active real memory KBytes sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds Memory is active when it is actually being used by a process running in memory This value should be less than or equal to memInUse A high value may indicate a system with insufficient memory numProcs 1 3 6 1 4 1 546 1 1 7 8 11 The number of processes table slots currently allocated and in use This number equals the number of currently running processes on the system A high value may indicate an overloaded system although higher values should be expected on servers and systems shared by many users 3 of 9 12 36 Data Source ReportPack Guide Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String numOpenFiles 1 3 6 1 4 1 546 1 1 7 8 12 The system wide total number of open files swapInUse 1 3 6 1 4 1 546 1 1 7 8 13 The amount of swap space KBytes currently in use by the system Swap space is used when memory is in short supply Typically entire processes are swapped out to secondary storage Comparing this value to totalSwap in Empire s kernelConfig group can help indicate if additional swap space is needed numSwitches 1 3 6 1 4 1 546 1 1 7 8 14 The total number of context switches that have occurred since the kernel was las
73. 1 56 4 5 249 41 0 02 156 portland 121 Accounting 1 56 4 4 249 83 0 02 166 torrance 1 89 Support 4 56 5 3 250 22 0 02 156 torrance 41 90 Marketing 41 10 6 9 140 13 0 04 877 torrance 188 Sales 1 10 7 8 140 08 0 04 875 Sanjose 4 20 Support 1 09 8 6 140 21 0 04 877 sanjose 1 40 Marketing 1 09 7 140 04 0 04 876 C A The PVCs with Greatest Change in Grade of Service GOS Score table shows the PVCs sorted by change in rank in ascending order based on the report day s GOS Score change Drastic increases in grade of service scores Report Descriptions 4 19 U o i x Yoday Aejay ewei4 v 4 Frame Relay ReportPack Service Level Management Report 4 20 suggest additional PVCs be watched while decreases document an effective remedy to a previously negative condition Table 4 3 describes how Grade of Service is calculated Device lincoln portland torrance torrance Sanjose sanjose torrance atlanta atlanta Interface DLCI 1 19 1 30 1 89 1 90 1 20 1 40 1 88 1 24 1 42 PVCs With Greatest Change In Grade Of Service GOS Score Accounting Sales Support Marketing Support Marketing Sales Marketing Marketing Description Wed Dec 22 1999 GOS Score 2 16 1 56 1 55 1 10 Previous GOS Score 2 15 1 55 1 56 1 09 1 10 GOS Change Rank GOS Rank Previous Ranl oo 4 OA Pw N 1 1 4 5 5 3 6 9 8 6 9 7 7 8 10 10 2 1 The Service Level Management report let CIOs CFOs network
74. 120 1 86 3105853 tulsa Cisco 2 48 29244859 1 08 19032892 boston Cisco 1 86 12021361 1 00 5195688 Double click on a router to update the content of following charts and graphs with performance data for the selected router Report Descriptions 6 11 o gt D fe c D D me e U o P x 6 Router ReportPack The Discards bar chart shows the total number of discards for the selected router for each sample taken on the report day Discards miami Acme s Main Office Thu Oct 21 1999 3 734 721 50 2 987 777 20 2 240 832 90 1 493 883 60 Total Discards 746 944 30 0 Thu 00 00 Thu 09 00 Thu 18 00 The Percent Discarded Packets bar chart shows the percentage of discards for the selected router for each sample taken on the report day Percent Discarded Packets miami Acme s Main Office Thu Oct 21 1999 5 60 4 20 Discards N p 1 40 00 00 06 00 12 00 18 00 00 00 6 12 Report Descriptions ReportPack Guide The Errors bar chart shows total number of errors for the selected router for each sample taken on the preceding day Errors miami Acme s Main Office Thu Oct 21 1999 4 133 937 50 3 307 150 00 2 480 362 50 1 653 575 00 Total Errors 826 787 50 0 Thu 00 00 Thu 09 00 Thu 18 00 o gt D fe c D D me fe U o P x The Percent Errored Packets bar chart shows the percentage of errors for the selected router for eac
75. 13 Top Ten 12 30 Index 5 Index Index 6 T targets selecting for snapshot report 1 27 Throughput definition of 2 6 Top Ten report description of 2 5 for Bay Router ReportPack 7 21 for Cisco Router ReportPack 8 25 for Frame Relay ReportPack 4 18 for LAN ReportPack 5 21 for RMON ReportPack 10 9 for Router ReportPack 6 15 for System ReportPack 12 30 for WAN ReportPack 5 21 TREND Database ReportPack 11 1 data source variables 11 6 installing 1 2 metrics for 11 11 report descriptions 11 1 Database Usage Summary 11 2 U uninstalling ReportPacks 1 25 Utilization definition of 2 6 V Volume definition of 2 6 WAN ReportPack data source variables 5 27 description of 5 1 Grade of Service calculation for 5 36 installing 1 2 metrics for 5 33 report descriptions 5 2 Capacity Planning 3 13 5 16 Executive Summary 5 2 Forecast 5 3 Hot Spots 5 17 Quick View Snapshot 5 6 Service Level Management 5 23 Top Ten 5 21
76. 15 minutes the CPU was not idle The agent continuously samples the CPU and updates this variable On multi processor machines this object represents the value for the first CPU only The Host Resources MIB hrProcessorTable breaks out the CPU load on a per processor basis This value is meaningless until at least two samples have been taken On NT systems this value is the average over all CPUs runQlen 1 3 6 1 4 1 546 1 1 7 8 4 1 of 9 12 34 Data Source ReportPack Guide Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The length of the scheduler s run queue sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds The system s load average is computed from the value of runQLen the load average is defined as the average runQLen value over some interval disk WaitNum 1 3 6 1 4 1 546 1 1 7 8 5 The number of jobs that are waiting on disk I O sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds A large value may indicate an overloaded system although larger values may be observed on file server systems pageWaitNum 1 3 6 1 4 1 546 1 1 7 8 6 The number of jobs that are waiting on page I O sampled over a fixed period of time The sampling interval is system dependent I
77. 1605 166739125 239835854 239031279 167047995 290036337 46660616 239692493 Previous Volume 238579410 289770889 240028759 166705624 238966889 238200269 166292234 289608971 46543182 239715997 Volume Volume Rank Previt Change Rank 4 on oom sp scores Volume and utilization for each interface is also shown Device houston boston sanluis miami sanjose miami sanjose tulsa houston cotati Interface LAN Interfaces With Highest Grade Of Service GOS Scores Description Stock Room Northeast Sales Corporate Offices Sales Sales hiktgs hiktgs Tulsa Research Center Accounting Cotati Sales Mon Dec 20 1999 GOS Score GOSRank Previous Rank Utilization 1 99 1 98 1 96 1 96 1 95 1 95 1 87 1 54 1 52 1 00 7 8 9 10 11 12 13 16 17 18 10 9 12 11 13 14 15 19 18 20 23 79 23 72 21 79 21 74 21 80 21 84 19 57 13 54 13 65 3 82 Volume 290898951 290036337 239692493 239041605 239835854 240488729 239031279 165739125 167047995 46660616 The LAN Interfaces with Greatest Change in Grade of Service Scores table shows the current and previous GOS scores for up to ten interfaces with the highest percentage of change It ranks the interfaces by highest GOS percentage of change Drastic increases in GOS scores changes identify interfaces that need to be monitored while GOS score decreases document an effective remedy to a previously n
78. 2 1 shows the Cisco Router Hot Spots report Summary Depending on the report type the summary information is presented as a table as shown in Figure 2 1 or a graph It highlights the macro issues associated with the report type For the Cisco Router Hot Spots report shown in Figure 2 1 the Problem Summary for the Day table identifies the Cisco routers listed in descending order by the greatest number of total exceptions that are exhibiting the most problematic performance In this example problematic performance is reported as exception conditions in the areas of CPU memory buffer utilization and buffer faults An excep tion condition occurs when the predefined threshold for a metric for example 95 CPU utilization is exceeded You can modify the default thresholds sup plied in the TREND ReportPacks to suit your environment N fe lt p Drill Down Analysis The drill down analysis components are presented as a combination of tables graphs line and area and charts bar stacked bar and pie They provide a detailed analysis of the individual metrics for the selected managed element over the appropriate time period preceding day or 6 week baseline period In Figure 2 1 tabbed drill down components enable you to examine CPU utilization memory utilization buffer utilization and buffer faults for the selected router over each hour of the preceding day These multiple related charts bring together the in
79. 2 20 i i E 4 10 et v ist o 2 00 00 06 00 12 00 18 00 x Forecast Report The Forecast report lists all routers that are within 90 days of reaching a discard threshold Those routers closest to the 90 day discard threshold are listed first The calculation for the threshold is from the baseline period The Forecast report comprises the following components The Estimated Days to Threshold summary table shows the number of days until the identified router exceeds the discard 5 Estimated Days to Threshold DTT Routers Within 90 Days Of Discard Threshold Router Current Projected DTT Discards Current Projected Current Projected Discards Discards 90 Days Pkts sec Pkts sec 90 Days Bytes sec Bytes sec 90 Days T m Report Descriptions 6 5 6 Router ReportPack Double click on a router to update the following components with performance data for the selected router The Percent Discards bar chart shows the percentage of discards for the selected router for each day of the baseline period Percent Discards sanjose Unassigned Fri Sep 10 1999 Thu Oct 21 1999 4 00 3 00 2 00 Percent Discards u Sep 10 Sep 24 Oct 08 The Discard Rate line graph shows the percentage of discards for the selected router for each day of the baseline period Discard Rate sanjose Unassigned Fri Sep 10 1999 Thu Oct 21 1999 1 367 014 00 1 025 260 50 683 507 00 Discards Second 341 753 50
80. 2 7 definition of 2 7 H Hot Spots report description of 2 4 for Bay Router ReportPack 7 18 for Cisco Router ReportPack 3 32 8 23 for Frame Relay ReportPack 4 15 for LAN ReportPack 5 17 for RMON ReportPack 10 2 for Router ReportPack 6 11 for System ReportPack 12 25 for WAN ReportPack 5 17 Installing ReportPacks 1 2 1 14 installing ReportPacks data displaying by customer 1 20 dependencies 1 2 L LAN ReportPack data source variables 5 27 dependencies 1 2 description of 5 1 Grade of Service calculation for 5 36 installing 1 2 metrics for 5 31 report descriptions 5 2 Capacity Planning 5 16 Executive Summary 3 2 5 2 Forecast 5 3 Hot Spots 5 17 Quick View Snapshot 5 6 Service Level Management 5 23 Top Ten 5 21 Latency definition of 2 7 metrics 2 6 availability 2 7 baseline 2 10 congestion 2 7 days to threshold forecast 2 8 calculation of 2 9 forecasting 2 8 forecasts calculation of 2 9 frequency distribution 2 8 2 14 grade of service 2 7 latency 2 7 ranking 2 13 throughput 2 6 utilization 2 6 volume 2 6 Index 3 Index Index 4 N Near Real Time QuickView report for Bay Router ReportPack 7 13 for Cisco Router ReportPack 8 15 for Frame Relay ReportPack 3 27 4 9 for System ReportPack 12 14 Near Real Time QuickView report for LAN ReportPack 5 13 for WAN ReportPack 5 13 Near Real Time Snapshot report for Bay Router ReportPack 7 13 for Cisco Router ReportPack 8 15 for Frame Relay ReportPa
81. 3 267 70 0 Thu 00 00 Thu 09 00 Thu 18 00 The Swap Capacity line graph shows the average percentage of swap capacity available for each hour of the report day for the specified system Percent Swap Capacity leyte 1 Thu Oct 21 12 00 AM 89 20 66 90 22 30 f 0 00 00 06 00 12 00 18 00 00 00 Report Descriptions ReportPack Guide The Total Swaps bar chart shows the total swaps for each hour of the report day for the specified system Total Swaps leyte 1 Thu Oct 21 12 00 AM 11 20 8 40 a LJ 3 560 a 3 E 2 80 Thu 00 00 Thu 09 00 Thu 18 00 The Exception Details table shows the time the sample was taken for each CPU and memory utilization page fault paging and swapping exceptions and the values for the offending conditions Table 12 6 describes how each metric is calculated Exception Details leyte 1 Thu Oct 21 04 15 AM Time Period Device Interface Description CPU Utilization Memory Page Faults Paging Swa Utilization Thu Oct 21 04 15 AM leyte 1 34 00 38 50 59480 00 31174 00 4 Thu Oct 21 04 30 AM leyte 1 38 00 44 57 69721 00 34468 00 4 Thu Oct 21 04 46 AM leyte 1 41 00 431 74868 00 36068 00 4 Thu Oct 21 05 00 AM leyte 1 47 00 49 88 79117 00 38724 00 4 Thu Oct 21 05 15 AM leyte 1 49 00 53 23 81705 00 41937 00 6 Thu Oct 21 05 30 AM leyte 1 48 00 59 88 90874 00 44029 00 6 Thu Oct 21 05 46 AM leyte 1 64 00 60 93 93370 00 48035 00 6 Thu Oct 21 06 00 AM leyte 1 57 00 65 46 98659 00 49329 00 6
82. 3 276658 sanjose 1 40 46 16 43 14 78 00 37 00 5 00 478 276347 torrance 1 88 44 96 43 36 77 00 38 00 4 98 431 274214 Sanjose 1 20 44 79 42 83 74 00 37 00 4 96 475 274056 orlando 1 30 36 09 35 30 79 00 36 00 3 89 3 91 214660 164077 76 00 36 00 lincoln Report Descriptions 4 9 gt TI o 3 D D o o lt D o fe 1 4 Frame Relay ReportPack Double click on a device in the table to update the content of the following drill down components with performance data for the device The PVC Utilization Based on CIR line graph shows the selected PVC s utilization CIR for each sample taken during the report day PYC Utilization Based on CIR orlando Accounting Server Thu Dec 23 09 00 AM Thu Dec 23 12 15 PM 196 80 kh 447 60 Ne 38 40 a W 49 20 0 09 00 09 30 10 00 10 30 11 00 11 30 12 00 Sent Util CIR Rec Util CIR The PVC Volume line graph shows volume for the selected PVC for each sample taken during the report day PY C Volume orlando Accounting Server Thu Dec 23 09 00 AM Thu Dec 23 12 15 PM 239 299 64K 179 474 73K 419 649 82K 59 824 91K 0 09 00 09 45 10 30 11 15 12 00 Sent Bytes Rec Bytes 4 10 Report Descriptions ReportPack Guide The PVC Utilization Based on Interface Speed line graph shows the selected PVC s utilization ifSpeed for each sample taken during the report day PVC Utilization Based on Interface Speed orlando
83. 31 Buffer Utilization chart see Buffer Utilization Drill Down on page 3 32 3 28 Near Real Time QuickView and Snapshot Reports ReportPack Guide Router Selection List Figure 3 26 shows each router and its performance statistics based on the data polled for the last 6 hours The statistic shown for each metric is the average of all the data collected for that statistic over the 6 hour period Router Selection List Select router to see near real time reports Router Slot CPU Utilization Memory Buffer Utilization Utilization sanjose 3 50 30 47 27 45 92 sanjose 5 50 03 46 01 46 63 Sanjose 4 37 88 37 82 37 67 sanluis 3 33 38 36 03 33 04 Figure 3 26 Near Real Time Router Selection List Bay Router Double click on a device in the table to update the content of the associated drill down components with performance data for the device swioday wo gt Ss lt S gt Ke J m z oO Near Real Time QuickView and Snapshot Reports 3 29 3 Analyzing TREND Reports CPU Utilization Drill Down Figure 3 27 shows the CPU Utilization line graph which displays the CPU utilization for each sample taken during the report period for the selected router CPU Utilization sanluis 1 Mon Dec 20 12 00 A Mon Dec 20 10 45 PM 99 00 79 20 59 40 Percent 39 60 19 80 0 00 00 06 00 12 00 18 00 Figure 3 27 Near Real Time CPU Utilization Drill Down Bay Router 3 30 Near Real Time QuickVi
84. 32 2 1 9 Number of octets received over this virtual circuit since it was created frCircuitLastTimeChange 1 3 6 1 2 1 10 32 2 1 11 The value of sysUpTime when last there was a change in the virtual circuit state 2 of 2 Basic Metric Calculations The performance metrics used in Frame Relay reports are based on the calculations listed in Table 4 2 Note that some charts modify these calculations Table 4 2 Frame Relay Metrics Metric Computed As In Utilization inoctets 8 delta_time CIR 100 or inoctets 8 delta_time ifS peed 100 1 of 2 4 26 Basic Metric Calculations ReportPack Guide Table 4 2 Frame Relay Metrics Metric Computed As inoctets is multiplied by 8 to give the number of incoming bits divided by delta_time to give bits per second divided by Committed Information Rate CIR which is collected as bits sec times 100 to yield a percentage Interface speed ifSpeed is used if CIR is not available Out Utilization outoctets 8 delta_time CIR 100 or Coutoctets 8 delta_time ifSpeed 100 outoctets is multiplied by 8 to give the number of outgoing bits divided by delta_time to give bits per second divided by Committed Information Rate CIR which is collected as bits sec times 100 to yield a percentage Interface speed ifSpeed is used if CIR is not available Congestion fecns becns outframes inframe
85. 5 16 The Capacity Planning report shows the most overutilized and underutilized interfaces based on a 90 day forecast value It allows the CIO and network manager to estimate utilization levels in 30 60 or 90 days Note For an in depth analysis of how to interpret the LAN or WAN Capacity Planning report and for chart illustrations see Capacity Planning Report on page 3 13 The Capacity Planning report comprises the following components The Overutilized Interfaces summary table shows the interfaces projected to be 40 or higher utilized for LAN interfaces or 80 or higher input or output utilization whichever is larger for WAN interfaces within the next 90 days ranked in descending order by the 90 day projected utilization value All calculations are based on 95th percentile utilization values and are displayed as Report Descriptions ReportPack Guide percentages in this table For more information about how utilization is computed see Table 5 2 The Underutilized Interfaces summary table shows the interfaces projected to be the least utilized within the next 90 days ranked in ascending order by the 90 day projected utilization value All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Note that the associated drill down components are the same for both underutilized and overutilized interfaces Double click on an interface in either table to update the
86. 522 93 4 1 273397 10 09 6 sanluis 3 48 27366425 47 6 7 27378710 53 5 Figure 3 16 Top Ten Largest Bay Router Throughput Changes The report shows the current day s discards and traffic in packets per second previous day s traffic in packets per second and current day s rank previous day s rank and change in rank from the previous day A large change in rank identifies a volatile resource and you should investigate the reasons for the high change in throughput Bay Routers with the Most Percentage of Discards Figure 3 17 shows the Bay Routers with the Most Percentage of Discards summary table which ranks Bay routers in descending order by the highest percentage of packet discards for the report day Packets second and bytes second for the selected Bay routers are also shown swioday wo gt lt S gt Ko J m z oO Bay Routers with Most Percentage of Discards Sun Dec 19 1999 Device Pct Pkts Pkts sec Rank Bytes sec Prior Rank Discarded sanjose 3 73 27538522 93 3 1182195575 79 3 sanluis 3 48 27366425 47 4 8534497 19 02 4 Figure 3 17 Top Ten Bay Routers with Most Percentage of Discards Top Ten 3 19 3 Analyzing TREND Reports Bay Routers with the Greatest Change in Percentage of Discards Figure 3 18 shows the Bay Routers with the Greatest Change in Percentage of Discards summary table which ranks the selected routers in ascending sequence by change in discard percentage from th
87. 6 Hot Spots 3 32 8 23 Near Real Time QuickView 8 15 Near Real Time Snapshot 8 15 QuickView 8 10 Service Level Management 8 26 Snapshot 8 10 Top Ten 8 25 Congestion definition of 2 7 D data displaying by customer 1 20 E elements selecting for snapshot report 1 27 Executive Summary report description of 2 4 for Bay Router ReportPack 7 2 for Cisco Router ReportPack 8 2 for Frame Relay ReportPack 4 2 for LAN ReportPack 3 2 5 2 for Router ReportPack 6 2 for System ReportPack 12 2 for WAN ReportPack 5 2 F Forecast report description of 2 4 for Bay Router ReportPack 7 6 for Cisco Router ReportPack 8 6 for Frame Relay ReportPack 3 7 4 6 for LAN ReportPack 5 3 for Router ReportPack 6 5 for WAN ReportPack 5 3 Forecast Filesystems report for System ReportPack 12 4 Forecasting definition of 2 8 Forecast Memory report for System ReportPack 12 8 Frame Relay ReportPack data source variables 4 25 dependencies 1 2 description of 4 1 Grade of Service calculation for 4 28 installing 1 2 metric for 4 26 report descriptions 4 2 Capacity Planning 4 12 Executive Summary 4 2 Forecast 3 7 4 6 Hot Spots 4 15 Near Real Time Quick View 3 27 4 9 Near Real Time Snapshot 3 27 4 9 QuickView 3 20 4 7 Snapshot 3 20 4 7 Top Ten 4 18 ReportPack Guide Frequency distribution definition of 2 8 sample calculation of 2 14 G Ganymede Pegasus ReportPack installing 1 2 GOS See Grade of Service Grade of Service
88. 6 80 60 10 33 40 Percent 16 70 o Sep 10 Sep14 Sep18 Sep22 Sep26 Sep30 OctO4 OctO8 Oct12 Oct16 Oct20 12 20 Report Descriptions ReportPack Guide Report Descriptions The Memory Utilization line graph shows the average memory utilization over each day of the baseline period for the selected system CPU Lti Memory Util Swap Util Page Faults Memory Utilization wasp Fri Sep 10 1999 Thu Oct 21 1999 74 40 55 80 37 20 Percent 18 60 o Sep10 Sep14 Sep18 Sep22 Sep26 Sep30 OctO4 OctO8 Oct12 Oct16 Oct20 The Swap Utilization line graph shows the average swap utilization over each day of the baseline period for the selected system CPU Util Memory Util Swap Util Page Faults Swap Utilization wasp Fri Sep 10 1999 Thu Oct 21 1999 66 80 50 10 B a z H 2 33 40 a a 16 70 o Sep10 Sep14 Sep18 Sep22 Sep26 Sep30 OctO4 OctO8 Oct12 Oct16 Oct20 12 21 _ N o lt 77 ca 3 D D O a U o 2 12 System ReportPack 12 22 The Page Faults bar chart shows the number of page faults for each day of the baseline period for the selected system CPU Util Memory Util Swap Util Page Faults Page Faults wasp Fri Sep 10 1999 Thu Oct 21 1999 11 484 948 00 8 598 336 00 6 732 224 00 2 866 112 00 Number of Page Faults oO Sep 10 Sep14 Sep18 Sep22 Sep26 Sep30 OctO4 OctO8 Oct12 Oct16 Oct20 The Underutilized Sys
89. 68 00 0 00 51 00 152 00 Double click on any system listed in the table to update the following components with performance data for the selection The Grade of Service stacked bar chart shows weighted values for memory swap and CPU utilization and queue length for the selected system over each hour on the report day Table 12 7 gives the GOS weighting factors Grade Of Service GOS leyte 1 Thu Oct 21 12 00 AM if niha Pii Excellent pjjma Thu 00 00 Thu 09 00 Thu 18 00 E Queue Length E Swap Utilization Memory Utilization E CPU Utilization 12 26 Report Descriptions ReportPack Guide Report Descriptions The CPU Utilization line graph shows the percentage of CPU utilization for each hour of the report day for the specified system CPU Utilization leyte 1 Thu Oct 21 12 00 AM 89 20 66 90 Percent 22 30 0 00 00 06 00 12 00 18 00 00 00 The Memory Utilization line graph shows the average percentage of memory utilization for each hour of the report day for the specified system Memory Utilization 89 20 66 90 Percent 22 30 0 00 00 06 00 12 00 18 00 00 00 12 27 _ N o lt 77 3 D D O a U o 2 12 System ReportPack 12 28 The Page Faults bar chart shows the number of page faults for each hour of the report day for the specified system Page Faults 173 070 80 129 803 10 86 535 40 Number of Page Faults 4
90. 8 20 Number of Frames Nov 11 Nov 25 DecO9 4 14 Report Descriptions ReportPack Guide Hot Spots Report The Hot Spots report provides the network manager and other members of the network management staff a list of PVCs experiencing utilization congestion or grade of service exceptions The top ten offenders ranked by total number of exceptions are presented in a summary table to assist the network management team in analyzing the problems Drill down charts display grade of service utilization and congestion charts as well as an exception detail table that shows the time of each exception and the value for the offending conditions as well as other categories The Hot Spots report comprises the following The Problem Summary for the Day table ranks the PVCs with the highest exceptions counts ranked by total number of exceptions input utilization output utilization congestion in descending order Problem Summary For The Day Frame Relay PVCs With Most Exceptions Wed Dec 22 1999 Device Element Description Congestion In Util Out Util Total Utilizatic Exceptions Exceptions Exceptions Exceptions Exceptio portland 1 30 Sales 0 00 81 00 81 00 162 00 162 00 torrance 1 89 Support 0 00 80 00 81 00 161 00 161 00 portland 1 21 Accounting 0 00 79 00 81 00 160 00 160 00 torrance 1 88 Sales 0 00 23 00 32 00 55 00 55 00 torance 1 90 Marketing 0 00 20 00 35 00 55 00 55 00 Sanjose 1 20 Support 0 00 23 00 31 00 54 00 54 00 sanjose 1 4
91. 8 82 56 81 87 81 18 85 87 E Figure 3 11 Capacity Planning Overutilized WAN Interfaces Summary By comparing this information with the interfaces listed in the Underutilized WAN Interfaces Summary table you can potentially improve performance by shifting the traffic load from overutilized to underutilized links If that is not possible the Overutilized WAN Interfaces Summary table provides advanced warning of interfaces where more bandwidth will soon be needed to prevent congestion from affecting end user service levels The Overutilized WAN Interfaces Summary table displays average Input and Output Utilization values for the baseline period for each interface and specific forecast Capacity Planning Report ReportPack Guide values for 30 60 and 90 days into the future The forecast values are the result of a least squares linear regression analysis performed on data samples from the baseline period TREND uses a rolling baseline For an explanation of how baselines are calculated see Baselines on page 2 10 The overutilized interfaces are sorted from the highest forecast 90 day utilization either input or output value to lowest Only interfaces projected to exceed the utilization threshold within 90 days are included in the table Depending on your network performance profile none of the WAN interfaces may meet this criteria or fewer than 10 interfaces may be listed To assist you in analyzing long term capacity trends
92. 9 Mon Dec 13 1999 260 00 195 00 130 00 65 00 Percent Utilization 0 Mon 06 Wed 08 Fri 10 Sun 12 Peak Segment Utilization Avg Segment Utilization The Collisions Dropped Frames and Events line graph shows the daily total collisions dropped events and dropped frames for the past six weeks if available for the interface selected from the Daily Segment Utilization table Collisions Dropped Frames and Events bluegrass ALL 3 Mon Dec 06 1999 Mon Dec 13 1999 0 01 Total 0 Mon 06 Wed 08 Fri 10 Sun 12 Segment Collisions Segment Dropped Events Segment Dropped Frames 10 6 Report Descriptions ReportPack Guide Report Descriptions The Segment Errors line graph shows the daily total errors by type for the past six weeks if available for the interface selected from the Daily Segment Utilization table Segment Errors bluegrass ALL 3 Mon Dec 06 1999 Mon Dec 13 1999 0 01 me 5 Z N me i xe e o Mon 06 Wed 08 Fri 10 Sun 12 aS 7 Segment Giants Segment CRC Brors Segment Fragments Segment Jabbers Segment Runts The Frame Size Distribution stacked bar chart shows the daily total number of frames and the distribution of frame size for the past six weeks 10 7 10 RMON2 ReportPack 10 8 if available for the interface selected from the Daily Segment Utilization table Frame Size Distribution bluegrass ALL 3 Mon Dec 06 19
93. 9 02 00 AM sanjose 3 82 21 93 81 94 09 Sun Dec 19 02 15 AM sanjose 3 82 21 90 17 98 95 Top Ten Report The Top Ten report lets the network manager know at a glance which Bay routers produce the greatest throughput and exhibit the poorest health The top ten Bay routers ranked by total throughput packets are listed by highest rank as well as highest change of rank Throughput and throughput change charts alert you to the routers greatest changes in throughput suggesting potential capacity overloading if increasing trends continue The throughput change report also may highlight the newly installed devices and technologies that are approaching a logical utilization threshold or are underutilized The Top Ten report comprises the following components Report Descriptions 7 21 N ius o lt D e 5 Po 1 xe e 1 7 Bay Router ReportPack The Busiest Bay Routers summary table shows up to ten Bay routers with the highest throughput packets for the report day Busiest Bay Routers Sun Dec 19 1999 Router Pct Discards Pkts sec Rank Bytes sec Prev Rank Sanjose 3 73 27538522 93 4 1182195575 79 6 sanluis 3 48 27366425 47 6 8534497 19 02 5 The Largest Bay Router Throughput Changes summary table ranks the selected routers in ascending sequence by change in throughput packets The report shows the current day s discards and traffic in packets per second previous day s traffic in packet
94. 99 Mon Dec 13 1999 1 155 072 00 866 304 00 577 536 00 Total Frames 288 768 00 0 MonO6 WedOS Frit10 Sun12 1024 1818 Bytes 512 1023 Bytes 256 511 Bytes M 128 255 Bytes 65 127 Bytes Ws Bytes Hourly QuickView Report The Hourly QuickView report lists each network interface with its corresponding average and peak utilization for the preceding day Associated charts show the hourly breakdown of segment utilization collisions errors and frame size distribution for the selected interface The Hourly QuickView report comprises the following components The Segment Utilization table lists the polled interfaces that you can select for further analysis in the associated charts Both the daily average utilization and the daily peak utilization are specified The interfaces are sorted in descending order based on the daily average utilization Report Descriptions ReportPack Guide Double click on an interface in the chart to update the content of the following components with performance data for the selected interface The Hourly Utilization line graph shows the average and peak utilization by hour for the interface selected from the Segment Utilization table The Collisions Dropped Frames and Events line graph shows the total collisions dropped events and dropped frames by hour for the interface selected from the Segment Utilization table The Segment Errors line graph shows the total errors by type by
95. 99 Thu Oct 21 1999 121 50 97 20 72 90 Percent 24 30 i Sep 09 Sep 23 Oct OF The CPU Utilization line graph shows the average CPU utilization for the selected filesystem for each day of the baseline period CPU Util Memory Util Swap Util Page Faults l CPU Utilization wasp C Fri Sep 10 1999 Thu Oct 21 1999 12 80 9 60 6 40 Percent 3 20 o Sep09 Sep13 Sep17 Sep21 Sep25 Sep29 OctO3 OctOF Oct11 Oct15 Oct 19 12 6 Report Descriptions ReportPack Guide The Memory Utilization line graph shows the average memory utilization for the selected filesystem for each day of the baseline period CPU Util Memory Util Swap Util Page Faults Memory Utilization wasp C Fri Sep 10 1999 Thu Oct 21 1999 30 80 23 10 F 15 40 a a 770 o Sep09 Sep13 Sep17 Sep21 Sep25 Sep29 OctO3 OctOF Oct11 Oct15 Oct19 The Swap Utilization line graph shows the average swap utilization for the selected filesystem for each day of the baseline period CPU Util Memory Util Swap Util Page Faults Swap Utilization wasp C Fri Sep 10 1999 Thu Oct 21 1999 20 10 a 13 40 A 8 6 70 oO Sep09 Sep13 Sep17 Sep21 Sep25 Sep29 OctO3 OctOF Oct11 Oct15 Oct 19 Report Descriptions 12 7 _ N o lt 77 ca 3 D D O a U o 2 12 System ReportPack The Page Faults bar chart shows the number of page faults for the selected files
96. Backward Explicit Congestion Notifications BECNs that have been received for that PVC Congestion notifications are indications of degrading service to your end users Congestion Sanjose Support Thu Nov 11 1999 Wed Dec 22 1999 2 471 60 1 853 70 1 235 80 617 90 Number of Frames 0 Now 11 Nov 25 DecO9 Frames w F ECNs Frames w BECNs Figure 3 10 Forecast Congestion Drill Down Frame Relay 3 12 Forecast Report ReportPack Guide Capacity Planning Report The Capacity Planning report details the most overutilized and the most underutilized interfaces This information can be used to Point out opportunities for balancing the load more evenly between interfaces to improve service levels without requiring additional investment Identify interfaces where more bandwidth will be required to meet user demands Expose interfaces where a lower less costly service would still meet user needs For example the following components comprise the Capacity Planning report for WAN Connectivity ReportPack Overutilized WAN Interfaces Summary summary table see Overutilized WAN Interfaces Summary on page 3 14 and the following drill down components Daily GOS Drill Down Chart see Daily Grade of Service Drill Down on page 3 15 Daily Utilization Drill Down Chart see Daily Utilization Drill Down on page 3 16 Daily Errors and Discards Drill Down Chart see Daily Errors amp
97. Dec 21 01 30 PM 142 183 20 106 637 40 71 091 60 Number of Packets 35 545 80 Mon 21 00 Tue 06 00 The Errors and Discards Percentages line graph shows the percent errors and percent discards for the selected interface for each sample taken during the report period Errors and Discards Percentages boston Northeast Sales Mon Dec 20 01 00 Phl Tue Dec 21 01 30 PM a D f iy 3 Percent of All Packets N b w o i Mon 21 00 Tue 06 00 Near Real Time Snapshot The Snapshot report comprises the following components Report Descriptions 5 15 5 LAN and WAN ReportPacks The Interface Selection List table lists the available interfaces Double click on an interface in the chart update the following components with performance data for the selected interface The Interface Average Traffic Rate line graph shows the average traffic rate for each sample period of the report day for the selected LAN interface The Interface Utilization line graph shows the average utilization for the percentage utilization for each sample period of the report day The Errors and Discards line graph shows the total errors and total discards for the selected interface for each sample period of the report day The Errors and Percent Discards Percentages line graph shows the percent errors and percent discards for the selected interface for each sample period of the report day Capacity Planning Report
98. Device Pct Pkts Discarded lincoln 3 73 miami 3 73 houston 2 58 cotati 2 52 tulsa 2 48 boston 1 86 Cisco Routers with Highest Discard Percentage Sun Dec 19 1999 Rank Prior Rank oun Oo aA N ow Oo a N Pkts sec 3392612371 27480781 14 11812231 66 4015947 52 35379888 04 37741747 27 Bytes sec 943368212 95 1182570806 32 972846001 10 174640564 83 947004189 12 1213756965 47 The Cisco Routers with Greatest Change in Discard Percentage table shows the Cisco routers with the greatest change in the average Discard Percentage score from the day preceding the report day to the report day The Cisco router with the greatest change in Discard Percentage rank is listed first Discarded Pkts Discarded Cisco Routers With Greatest Change in Discard Percentage Sun Dec 19 1999 Change Rank Rank Prior Rank Pkts sec Device Pct Pkts cotati 2 52 lincoln 3 73 boston 1 86 miami 3 73 houston 2 58 tulsa 2 48 Previous Pct 2 53 3 74 1 85 3 73 2 58 2 48 o Ny aA U N 4 aN O Service Level Management Report 6 4015947 52 33926123 71 37741747 27 27480781 14 11812231 66 au an O 36379888 04 Bytes sec 174640564 83 443368212 95 1213756965 47 118257080632 972845001 10 947004189 12 The Service Level Management report lets CIOs CFOs network managers users and customers know how their Cisco routers are performing against contracted service levels It reports the ten routers
99. Discards Drill Down on page 3 17 swioday Underutilized WAN Interfaces Summary table and the following drill down components Daily GOS Drill Down Chart i gt lt Ss Ko J m z oO Daily Utilization Drill Down Chart Daily Errors and Discards Drill Down Chart Capacity Planning Report 3 13 3 Analyzing TREND Reports Note The following sections describe overutilized interfaces only as the analysis information for underutilized interfaces is the same The following sections describe these charts in more detail For more information about the WAN Connectivity ReportPack see LAN and WAN ReportPacks on page 5 1 Overutilized WAN Interfaces Summary Figure 3 11 shows the Overutilized WAN Interfaces summary table which lists the top 10 interfaces projected to exceed the utilization threshold in the next 90 days Overutilized WAN Interfaces Projected To Exceed 80 Utilization Within 90 Days All Values Based On 95th Percentile P95 Device Element Description Current P95 Projectedin Projectedin Projectedin Current P95 In Util Util 30 Days Util 60 Days Util 90 Days Out Util eae eee boston 3 WAN Link Boston San Jose 86 09 94 82 102 63 110 44 87 08 Sanjose 3 WAN Link New Yor 55 48 66 94 77 43 87 91 66 54 miami 3 WAN Link New Yor 55 47 67 25 77 70 88 14 67 34 sanluis 4 WAN Link Atlanta Portland 84 57 81 87 80 91 79 96 85 38 tulsa 3 WAN Link Tulsa Lincoln II 85 6
100. LL 0 21 0 80 0 00 0 00 Double click on an interface in the table to update the content of the following components with performance data for the selected interface The RMON Ethernet Utilization line graph shows the daily average 95th percentile and peak utilization for the past 30 days if available for the interface selected in the Daily RMON Ethernet Statistics table RMON Ethernet Utilization bluegrass 1_ALL Tue Sep 21 1999 Wed Oct 20 1999 69 50 55 60 41 70 27 80 Percent Utilization 13 90 Sep 28 Oct 05 Oct 12 Oct 19 Average Utilization P96 Utilization Peak Utilization 9 4 Report Descriptions ReportPack Guide The RMON Ethernet Collisions line graph shows the daily average and peak percent collisions for the past 30 days if available for the interface selected in the Daily RMON Ethernet Statistics table RMON Ethernet Collisions bluegrass 1_ALL Tue Sep 21 1999 Wed Oct 20 1999 o Pe fe Z poj fe U o Q 0 20 0 10 Percent Collisions Sep 28 Oct 05 Oct 12 Oct 19 Awg Pet Collisions Peak Pet Collisions Monthly RMON Ethernet The Monthly RMON Ethernet report lists each Ethernet interface with its corre sponding average and peak utilization and average and peak percent collisions for the preceding month It comprises the following components The Monthly RMON Ethernet Statistics table provides a list of all Ethernet interfaces with their mon
101. Large Buffer Utilization Medium Buffer Utilization Small Buffer Utilization The Daily Total Volume chart shows the total number of bytes volume for the selected Cisco router for each day of the baseline period For more information about how Volume is computed see Table 7 2 Total Bytes Discards Daily Total Volume tulsa Desktalk Support 4 hon Now 08 1999 Sun Dec 19 1999 57 917 577 20 43 438 182 90 28 958 783 60 Megabytes 14 479 394 30 0 Nov 08 Nov 22 Dec 06 8 22 Report Descriptions ReportPack Guide The Daily Percent Discards chart shows the percentage of discards for the selected Cisco router for each day of the baseline period For more information about how Discards is computed see Table 7 2 Total Bytes Discards Daily Percent Discards tulsa Desktalk Support 4 Mon Now 08 1999 Sun Dec 19 1999 3 00 Percent 0 Nov 08 Nov 22 Dec 06 Hot Spots Report This report lists Cisco routers that have exceeded CPU Utilization Memory Utiliza tion Buffer Utilization or Buffer Faults threshold conditions during the report day Offending Cisco routers are ranked by total number of exceptions in the Problem Summary table Each exception is linked to drill down charts that display informa tion for the following hourly Grade of Service CPU utilization buffer utilization memory utilization percentage of discards total buffer faults and exception detail Note For
102. Now22 Now29 DecO6 Dec13 Availability Service Level The Hourly Availability line graph shows the average hourly availability and service level for the selected Cisco router for each sample for the preceding day By default the service level is set at 99 5 Daily Hourly Hourly Availability miami Sun Dec 19 12 00 AM Mon Dec 20 12 00 AM 100 00 80 00 60 00 Percent 02 00 06 00 08 00 11 00 14 00 17 00 20 00 23 00 Availability Service Level 8 28 Report Descriptions ReportPack Guide The Response Time table lists the Cisco routers with the highest average response times for the day For more information about how response time is calculated see Response Time on page 3 45 Response Time Routers With Highest Response Time Sun Dec 19 1999 Device Response Time miami 500 00 boston 479 52 cotati 37 75 tulsa 362 31 lincoln 272 95 Double click on a Cisco router to update the content of the following components with performance data for the selected router The Daily Response Time line graph shows the average response time for the selected Cisco router for each day of the baseline period Daily l Hourly l Daily Response Time houston Acme s Support Office Mon Nov 08 1999 Sun Dec 19 1999 685 50 548 40 411 30 274 20 milliseconds 137 10 Nov 15 Novw22 Nov29 DecOB Dec13 D D S e U o i x J9 NOY ODSID 8 Repor
103. ON version 4 0 Check that an earlier version is installed dependency RMON version 3 5 3 Using AutoPilot add the UPGRADE_RMON ReportPack For more informa tion about running AutoPilot see Running Autopilot on page 1 10 D U D Q 2 gt 2 3 D 1 Note Before you can add the RMON2 ReportPack you must run an upgrade procedure for the V3 6 binary set to operate correctly For more information about running the upgrade procedure see Running the RMON2 ReportPack Upgrade Procedure on page 1 7 Running the RMON2 ReportPack Upgrade Procedure For the RMON2 ReportPack only you must run an upgrade procedure for the V3 6 binary set to operate correctly The following bug is being fixed by this upgrade Incorrect translation of DECNET and AIX addresses in host and matrix RMON2 tables The rmon2_341_upgrade exe program performs following changes Adds a new column address_type to the dsi_protocol table Populates the address_type column in the dsi_protocol table Creates and populates the dsi_version_info table o Updates the dsi_table_key nl_srcaddress and nl_dstaddress columns in RMON2 key tables Installing ReportPacks 1 7 1 ReportPack Administration Note The rmon2_341_upgrade exe program does not change the table_keyxxx column in data tables the changes are made only to the column in Key Tables Some Key Tables Ghost Key Tables
104. Pack The Filesystem Utilization line graph shows the selected system s filesystem utilization for each sample taken during the report period Filesystem Utilization wasp Thu Oct 21 06 00 PM 98 00 73 50 49 00 wasp C wasp E 24 50 a 18 00 00 00 06 00 12 00 Capacity Planning Filesystems Report The Capacity Planning Filesystems report shows the most overutilized and un derutilized systems that are within 90 days of reaching a disk space utilization threshold It allows the CIO and network manager to estimate utilization levels in 90 days You can double click on any interface in the Overutilized Systems or Un derutilized Systems summary tables to display corresponding drill down charts that detail daily data for Grade of Service page faults and CPU memory and swap uti lization for the baseline period It shows summaries of the filesystems that are fore casted to exceed 90 of the space used on the filesystem in the next 90 days for overutilized and less then 30 of the space used on the filesystem in the next 90 days for underutilized The Capacity Planning Filesystems report comprises the following components The Overutilized System Filesystem summary table shows the filesystems that are projected to be overutilized within the next 90 days They are ranked in descending order by the 90 day projected filesystem utilization value The filesystems using the most projected space within the next 90 days are sho
105. ReportPack installation Using the property import feature Selecting deferred targets or elements for a snapshot report 1 ReportPack Administration ReportPack Dependencies Some ReportPacks use reports from other ReportPacks Therefore ensure that you install the required ReportPack s if any for the ReportPack you are using Table 1 1 lists the dependencies for each ReportPack Table 1 1 ReportPack Dependencies ReportPack Dependencies Frame Relay None LAN WAN None Router LAN WAN Bay Router Router LAN WAN Cisco Router Router LAN WAN RMON None RMON2 None TREND Database None System None Installing ReportPacks The V4 0 Bundled ReportPack suite comprises the following ReportPacks Frame Relay LAN_WAN Connectivity Router 1 2 ReportPack Dependencies ReportPack Guide Bay Router Cisco Router RMON RMON2 TREND_Database replaces the Database ReportPack D U D Q 2 gt 2 3 D 1 gt System When you installed TREND V3 6 the installation program placed the ReportPacks sources in the DPIPE_HOME packages directory For more information about in stalling the ReportPacks from those sources see Running Autopilot on page 1 10 Note In this chapter DPIPE_HOME represents the directory into which TREND has been installed By default this is the filesystem with the mos
106. Sun Dec 19 1999 Nov 08 Nov 22 Dec 06 Di buffer Faults E Buffer utilization Dtemory Utilization HE CPU Utilization Each metric shows the 95th percentile of the value with the Grade of Ser vice GOS weight applied for the selected Cisco router for each day of the baseline period Table 8 3 describes how Grade of Service is calculated Report Descriptions 8 7 D D xe e U o i x JOINOY ODSID 8 8 Cisco Router ReportPack The Daily CPU Utilization line graph shows the weighted average and 95th percentile CPU utilization for the selected Cisco router for each day of the baseline period Daily CPU Utilization tulsa Desktalk Support 4 tion Nov 08 1999 Sun Dec 19 1999 123 20 92 40 61 60 Percent 30 80 0 Nov 08 Nov 22 Dec 06 P95 CPU Utilization Awg CPU Utilization The Daily Memory Utilization line graph shows the weighted average and 95th percentile memory utilization for the selected Cisco router for each day of the baseline period Memory Buffer Daily Memory Utilization tulsa Desktalk Support 4 hon Now 08 1999 Sun Dec 19 1999 108 00 72 00 Percent 36 00 Nov 08 Nov 22 Dec 06 P95 Memory Utilization Awg Memory Utilization 8 8 Report Descriptions ReportPack Guide The Daily Buffer Utilization line graph shows the weighted average of the buffer utilization samples collected for the selected Cisco router for each day of the bas
107. T is more than 1 000 days it is assigned a null value which is displayed as N A in the reports If a metric has already exceeded a threshold value and is still rising the DTT will be a negative number indicating the number of days past the threshold If a metric is declining whether it is currently over threshold or not the DTT forecast is always null For the Frame Relay Forecast report the forecast metrics are Input Utilization and Output Utilization The Frame Relay ReportPack utilizes information collected from the RFC 1315 Frame Relay MIB TREND performs a least squares linear regression analysis on data samples from the baseline period to forecast when either Utilization metric will reach the threshold The default threshold values are predefined by TREND when the Frame Relay ReportPack is installed Forecast Report ReportPack Guide To provide a higher quality forecast for planning purposes TREND bases the linear regression analysis on the 95th percentile of utilization samples taken during the day rather than on a simple average The g5th percentile is the value below which 95 of the samples fall The g5th percentile better characterizes the shape of the data by taking into account the variations or spikes in network utilization that occur during the day rather than smoothing the data out by averaging By using 95th percentile in stead of average utilization values TREND forecasts provide better advanced warn ing and bandwidth
108. TREND ReportPack Guide TREND 3 6 ReportPacks 4 0 DeskTalk Systems Inc DeskTalk Systems Inc makes no representation or warranties with respect to this manual or except as specifically stated in the applicable user agreement or warranty notice with respect to any hardware firmware or software described in this manual DeskTalk Systems Inc spe cifically disclaims any expressed or implied warranties of merchantability title or fitness for a particular purpose Furthermore DeskTalk Systems Inc reserves the right to make revi sions or changes to any and all parts of the manual hardware firmware or software at any time without obligation to notify any person or entity of the changes DeskTalk DeskTalk Systems the DeskTalk logo and TREND product names are registered trademarks or trademarks of DeskTalk Systems Inc Other brand and product names are reg istered trademarks or trademarks of their respective holders Copyright 1992 2000 DeskTalk Systems Inc All rights reserved No part of this publication may be reproduced photocopied stored in a retrieval system transmitted or translated into any language without the prior written permis sion of DeskTalk Systems Inc TREND ReportPack Guide Publication No RPR361 2000JUN16 DeskTalk Systems Inc 19191 S Vermont Avenue Ninth Floor Torrance CA 90502 310 630 1000 TREND Contents FIQUICS 2 schec eee eed dade Sowa et ew deeded eoene eres xi Tables a eae xi
109. Utilization line graph shows the daily average 95th percentile and peak utilization for the past 30 days if available for the interface selected in the Daily RMON Token Ring Statistics table RMON Token Ring Utilization bluegrass 3_ALL 64000 Tue Sep 14 1999 Wed Oct 13 1999 Percent Utilization a o Sep 21 Sep 28 Oct 05 Oct 12 Awg Pet Util 96th Pet Util Peak Pet Util 9 8 Report Descriptions ReportPack Guide The RMON Token Ring Beacons line graph shows the daily average and peak beacons per second for the past 30 days if available for the interface selected in the Daily RMON Token Ring Statistics table RMON Token Ring Beacons bluegrass 3_ALL 64000 Tue Sep 14 1999 Wed Oct 13 1999 0 01 Beacons per Second Sep 21 Sep 28 Oct 05 Oct 12 Awg Beacons Sec Peak Beacons Sec Monthly RMON Token Ring The Monthly RMON Token Ring report comprises the following components The Monthly RMON Token Ring Statistics table provides a list of all Token Ring interfaces with their monthly average and peak utilization average Report Descriptions 9 9 o D fe Z pej 3 fe U o Q 9 RMON ReportPack beacons per second and total beacons The interfaces are sorted in descending order based on monthly peak utilization RMON Probe tms east desktalk com fp2000 east desktalk com bluegrass bluegrass bluegrass bluegrass bluegrass bluegrass Monthly RMON Tok
110. VCs closest to the 90 day threshold are listed first The forecast data is generated using the daily 95th percentile values for in and out utilization from the daily exception table The forecast is based on the data from the previous 42 days The report uses only the 90 day forecast for a constraint However the table will display 30 60 and 90 day forecasts Drill down charts display Grade of Service Utilization and Congestion information for that particular circuit over the baseline period by default 6 weeks Note For an in depth analysis of how to interpret the Frame Relay Forecast report and for chart illustrations see Forecast Report on page 3 7 Report Descriptions ReportPack Guide The Forecast report comprises the following charts The Estimated Days to Threshold table ranks the PVCs in ascending order that will reach a 80 In Utilization and Out Utilization threshold within 90 days Double click on a device to update the content of the following drill down components with performance data for the device The Grade of Service stacked bar chart shows Grade of Service GOS scores for In Utilization Out Utilization and Congestion for the selected PVC for each day of the baseline period Table 4 3 describes how Grade of Service is calculated The Utilization line graph shows In Utilization and Out Utilization computed using both CIR and ifSpeed for the selected PVC for each sample period on the repo
111. able System When the System ReportPack is installed the Availability SLA metric is set to 99 5 by default The Availability chart shows the 10 systems with the lowest availability Daily Availability Drill Down Figure 3 39 shows the Daily Availability line graph You can double click any system in the Availability table to update this graph with the average daily 3 42 Service Level Management Report ReportPack Guide availability over the report period previous 42 days for analysis and investigation of availability changes Daily Hourly l Daily Availability wasp 1 Fri Sep 10 1999 Thu Oct 21 1999 100 00 80 00 60 00 Availability 40 00 Service Level 20 00 0 Sep 10 Sep 24 OctOs Figure 3 39 Service Level Daily Availability Drill Down System Availability is a critical metric in most SLAs In principle availability is simply the percent of time that the system is up during any given time period Accessible data to measure availability and the quality of that data differ from device to device Hourly Availability Drill Down Figure 3 40 shows the Hourly Availability line graph You can double click any system in the Availability table to update this graph with the average hourly Service Level Management Report 3 43 swioday wo gt lt S gt Ko J m z s 3 Analyzing TREND Reports availability over the report period previous day for analysis and investigation
112. ach comprise 20 of the final GOS score and Congestion comprises the remaining 60 By assigning a grade and corresponding description for example Excellent Good Warning Critical to each time interval TREND provides an aggregate view of the stability or changes in overall health for the element Utilization Drill Down Figure 3 9 shows the Utilization line graph which presents input and output utilization values both as a percent of CIR Utilization Sanjose Support Thu Now 11 1999 Wed Dec 22 1999 Si gt 104 70 J lt ON 67 80 of 3 oF 5 Q a 33 90 oO Bs m 0 Z Now 11 Nov 25 Dec 09 oO P95 In Util CIR P95 Out Util CIR P95 In Util Ifspeed P95 Out Util Ifspeed Figure 3 9 Forecast Utilization Drill Down Frame Relay If a CIR has been defined this graph helps identify PVCs with utilization exceeding the CIR which can cause congestion and frames to be discarded By reviewing the Forecast Report 3 11 3 Analyzing TREND Reports changes and trends in utilization over the baseline period you can immediately understand whether utilization that exceeds CIR represents is a brief period of heavy traffic that is not likely to be repeated or a long standing situation of steadily growing traffic Congestion Drill Down Figure 3 10 shows the Congestion line graph which itemizes the number of Forward Explicit Congestion Notifications FECNs and
113. age Memory totalmemoryused The ratio of total memory in bytes Utilization etotalmemorymax 100 used to total memory available in bytes is multiplied by 100 to yield a percentage Buffer totalbuffersused The ratio of total number of buffers Utilization totalbuffersmax 100 used to the total number of buffers available is multiplied by 100 to yield a percentage Volume inoctets outoctets inoctets and outoctets bytes are totaled for the selected Bay router interface for the specified aggregation period The source of this metric is the data collected for the LAN WAN ReportPacks 1 of 2 7 28 Basic Metric Calculations ReportPack Guide Table 7 2 Bay Router Metrics Metric Computed As How Computed Errors ifouterrors This metric is essentially the ratio of ifinerrors the sum of inbound and outbound inunknownprotos packets containing errors and those ifoutucastpkts that are discarded because they come ifoutnucastpkts from an unknown source to the total inunknownprotos packet traffic inbound and ifinerrors outbound at the interface plus 1 In ifinucastpkts this computation a negligible ifinnucastpkts amount i e 1 packet is added to the ifindiscards 1 100 0 divisor to prevent a potential divide by zero condition The result is multiplied by 100 to yield a percentage The source of this metric is the data collected for the LAN WAN ReportPack
114. all Bay routers over each day of the baseline period up to and including the report day N Hourly Baseline Daily 3 Daily Volume All Bay Routers A Sun Nov 21 1999 Sat Jan 01 2000 z2 o 34 80 za o 26 10 Q 1 Gigabytes i B 8 70 Now 21 Dec 05 Dec 19 Hourly Grade of Service stacked bar chart provides an overview of Bay router performance by supplying weighted values for CPU memory and Report Descriptions 7 3 7 Bay Router ReportPack buffer utilization for each hour of the report day Table 7 3 describes the GOS weighting factors used in this chart Hourly Grade GOS All Bay Routers Sat Jan 01 12 00 AM Sun Jan 02 12 00 AM Critical Warning Good Excellent Sat 00 00 Sat 09 00 Sat 18 00 Bitemory Utilization M CPU Utilization E Buffer utilization The Hourly CPU Utilization line graph shows average and 95th percentile CPU utilization for each hour of the report day Hourly CPU Utilization All Bay Routers Sat Jan 01 12 00 Ai Sun Jan 02 12 00 Aut 130 40 97 80 65 20 32 60 0 00 00 06 00 12 00 18 00 00 00 P95 CPU Utilization Awg CPU Utilization 7 4 Report Descriptions ReportPack Guide The Hourly Buffer Utilization line graph shows average and 95th percentile buffer utilization for each hour of the report day Hourly Buffer Utilization All Bay Routers Sat Jan 01 12 00 AM Sun Jan 02 12 00 AM 130 00 97 50 65 00 32 50 0 00 00 06 00
115. an in depth analysis of how to interpret the Cisco Router Hot Spots report and for chart illustrations see Hot Spots Report on page 3 32 The Hot Spots report comprises the following charts D o S e U o i x JOINOY ODSID 8 Report Descriptions 8 23 8 Cisco Router ReportPack 8 24 The Problem Summary for the Day table ranks the Cisco routers with the highest total CPU utilization buffer utilization memory utilization and buffer fault exceptions counts by total number of exceptions cpuutil memutil bufutil bufferfaults in descending order Double click on a router to update the content of the following components with performance data for the selected router The Hourly Grade of Service stacked bar chart shows the Grade of Service GOS ratings for CPU Utilization Physical Memory Utilization Buffer Utilization and Buffer Errors for the selected Cisco router for each sample taken on the report day for the selected Cisco router Table 8 3 defines the GOS weighting for each of these metrics The CPU Utilization area graph shows CPU Utilization for the selected Cisco router for each sample period on the report day The Buffer Utilization line graph shows Buffer Utilization for the selected Cisco router for each sample taken on the report day for each buffer size There are five buffer sizes huge big large medium and small The Memory Utilization area graph shows Memory Utilization for
116. anagement Each report generally shows the same type of information for all ReportPacks For example the Forecast report for the Router ReportPack forecasts router resources for the next 30 60 and 90 days and identifies the days to threshold DTT for selected router resources The Forecast report for the Frame Relay ReportPack provides the same information for selected Frame Relay PVCs 3 1 3 Analyzing TREND Reports The way metrics are computed differ from one ReportPack to the next For example utilization for the Frame Relay ReportPack is given in terms of PVC bandwidth Conversely utilization for the Cisco Routers ReportPack is given in terms of CPU memory and buffer consumption This chapter analyzes selected reports from each ReportPack For example the Executive Summary report from the LAN ReportPack and the Forecast report from the Frame Relay ReportPack are analyzed in detail The intent of this approach is to give you guidelines for interpreting the information provided in any one report type If you apply the detailed explanation of a report contained in this chapter along with the ReportPack specific metrics in subsequent chapters you should be able to interpret the reports in any ReportPack Executive Summary Report 3 2 The Executive Summary report provides high level graphs aggregating key metrics for all managed elements within a technology for example all LAN interfaces all Frame Relay PVCs all WAN Inter
117. and Snapshot Reports syoequodey The Near Real Time QuickView and Near Real Time Snapshot reports provide performance statistics up to the last SNMP poll The report does not rely on nightly summaries therefore it provides instant reporting on collected data for individual LAN and WAN interfaces o gt Z 2 gt a The Snapshot report provides the same information as the QuickView report for individual interfaces that you select from a pick list for inclusion in the report For more information about selecting a interface for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Near Real Time QuickView The QuickView report comprises the following components The Interface Selection List table lists the available interfaces and its performance statistics based on the data polled for the last 6 hours The statistic shown for each metric is the average of all the data collected for that statistic over the 6 hour period Interface Selection List Select interface to see near real time reports Device Interface Speed Aug Utilization Peak AugPct Avg Pct Errors Description Utilization Discards houston 3 10000000 11 99 12 08 2 05 0 00 Stock Room miami 2 9000000 10 65 11 10 2 58 1 84 hiktgs sanjose 1 9000000 10 48 10 94 2 56 1 82 Sales miami 1 9000000 10 29 10 76 2 59 1 91 Sales sanluis 2 9000000 10 27 10 70 2 64 1 91 Corporate Offices sanjose 2 100000
118. andiego I sanjose 2 sanluis D santaclara EHC santarosa HOS Beer I Use selected Targets Elements for all data definitions ce Back Next Finish Cancel 2 Select the target or element you want to view from the Available Targets Elements box and click the arrow to move it to the Selected Targets Elements box 3 Select any other elements you want to view or if you are finished click Next Selecting Deferred Targets or Elements for a Snapshot Report 1 27 1 ReportPack Administration The Snapshot report is displayed with performance data for the selected target or element 1 28 Selecting Deferred Targets or Elements for a Snapshot Report TREND 2 Overview This chapter Describes the uniform layout of the reports in all TREND ReportPacks see Report Layout on page 2 2 Lists the eight reports that are common to most ReportPacks and briefly describes the content and purpose of each report see Standard Reports on page 2 4 Identifies the metrics volume throughput and so on that are reported for each enterprise technology area Frame Relay Circuits Routers and so on and gives an overview of how each metric is calculated see Metric Classes on page 2 6 Explains the concepts of forecasting baselining ranking and frequency distribution as they are applied in TREND reports see Forecasting on page 2 8 Baselines
119. ange 40 and over 10 and over 10 and over Table 5 5 lists how the metrics are weighted in WAN Grade of Service GOS charts Table 5 5 WAN GOS Calculations Total Total Input Output Percent Percent GOS Calculation Utilization Utilization Errors Discards Weighting 30 30 20 20 Excellent Score Range 0 70 0 70 under 5 under 5 Good Score Range 70 80 70 80 5 7 5 7 Warning Score Range 80 90 80 90 7 10 7 10 Critical Score Range gt 90 gt 90 over 10 over 10 For more information about the Grade of Service metric see Table 2 2 on page 2 6 For more information about Grade of Service as it applies to each type of report see Analyzing TREND Reports on page 3 1 5 36 Grade of Service Calculation TREND 6 Router ReportPack The Router ReportPack provides comprehensive data collection and reporting to support management of router networks It includes compound reports for capacity planning service level reporting forecasting and troubleshooting It also provides an executive summary on the performance of all routers combined and detailed reports on the performance of individual routers o a fe c a D me e U o gt The Router ReportPack answers questions including Do the router networks provide an acceptable level of service to customers or end users Are there overloaded routers If so
120. ant implementation of ASN 1 and BER must be able to generate and recognize this value 50f5 Basic Metric Calculations Table 5 2 lists the calculations for the performance metrics used in the LAN reports Note that some charts modify these calculations Table 5 2 LAN Metrics Metric Computed As Utilization ifoutoctets ifinoctets 8 delta_time ifspeed 1 100 0 This metric is the sum of inoctets and outoctets multiplied by 8 to give the number of bits which is divided by total speed to give bits per second Total bits per second is divided by ifspeed which is collected as bits per second plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Basic Metric Calculations 5 31 5 LAN and WAN ReportPacks Table 5 2 LAN Metrics Metric Computed As Errors ifouterrors ifinerrors inunknownprotos ifoutucastpkts ifoutnucastpktst inunknownprotos ifinerrors ifinucastpkts ifinnucastpkts ifindiscards 1 100 0 This metric is the ratio of the sum of inbound and outbound packets containing errors and those that are discarded because they come from an unknown source to the total packet traffic inbound and outbound at the interface plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Discards ifindiscards ifoutdiscards ifinucastpkts ifinnucastpkts ifindiscards ifinerrorst
121. appearing in the QuickView report Daily QuickView Report The Daily QuickView report lists each network interface with its corresponding average and peak utilization for the preceding day The associated charts show the history of daily segment utilization collisions errors and frame size distribution for the selected interface D e Z N D fe V D Q The Daily QuickView report comprises the following components The Daily Segment Utilization table lists of all polled interfaces that you can select for further analysis in the associated charts Both the daily average utilization and the daily peak utilization are specified The interfaces are sorted in descending order based on the daily average utilization Daily Segment Utilization Sorted by Average Utilization Mon Dec 13 1999 RMON Probe LF Type Speed Average Peak Utilization Utilization fp2000 east desttalk com 1_ALL ethernetcsmacd 10000000 00 1 73 15 33 bluegrass 1_LHS_GEN ethermet csmacd 10000000 00 0 00 0 00 Double click on an interface in the chart to update the content of the following components with performance data for the selected interface Report Descriptions 10 5 10 RMON2 ReportPack The Daily Utilization line graph shows the daily average and peak utilization for the past six weeks if available for the interface selected from the Daily Segment Utilization table Daily Utilization bluegrass ALL 3 Mon Dec 06 199
122. are affected by the upgrade Note For UNIX systems DeskTalk recommends that you back up the Ghost Key Tables before running the rmon2_341_upgrade exe program For more information about backing up the Ghost Key Tables and restoring key tables see Chapter 5 of the TREND Installation Guide The following Key Tables Ghost Key Tables are affected by upgrade Gmon2_nlhostdata_ applicable to users who collected the RMON2 NL HOST table with the 3 41 version of the collector Gmon2_alhostdata_ applicable to users who collected the RMON2 AL HOST table with the 3 41 version of the collector Gmon2_nlmatrixdata_ applicable to users who collected the RMON2 NL MATRIX table with the 3 41 version of the collector Gmon2_almatrixdata_ applicable to users who collected the RMON2 AL HOST table with the 3 41 version of the collector Note Ensure that all the rmon2 collections are turned off while the upgrade procedure is running To run the RMON2 upgrade procedure 1 Run the following program SDPIPE_HOME bin rmon2_341_upgrade exe 1 8 Installing ReportPacks ReportPack Guide 2 Using AutoPilot add the UPGRADE_RMON2 ReportPack For more infor mation about running AutoPilot see Running Autopilot on page 1 10 D U D Q 2 gt 2 3 D 1 Troubleshooting the RMON2 Upgrade Procedure If an error occurs during the upgrade procedure examine
123. are there under utilized routers that can help balance the load Are performance problems due to high levels of discards errors excessive utilization or other causes What are the top ten routers contributing to the poor health of the network Which routers are likely to degrade service in the near future unless you take preventive action 6 1 6 Router ReportPack Each report consists of a combination of charts graphs and tables To produce the reports the Router ReportPack discovers all routers collects their volume and discard data and then processes the data automatically Report Descriptions The following sections described each report in the Router ReportPack in detail Executive Summary Report The Executive Summary consists of charts and graphs that show executives and high level managers how corporate router networks are performing in terms of volume Note The TRENDweb Administrator can configure the Router ReportPack installation so that the Executive Summary reports will only display data for devices belonging to a specific customer In Figure 3 1 the report displays information for only the devices belonging to the Acme company For more information about this feature see Displaying Data for Specific Customer Devices in Reports on page 1 20 The Executive Summary report comprises the following components 6 2 Report Descriptions ReportPack Guide The Router Inventory summary
124. ati 437 75 sanjose 386 11 tulsa 362 31 lincoln 272 95 Double click a router in the table to update the following graphs with performance data for the selected router Report Descriptions 6 19 o gt D e c a D D gej e V o ie x 6 Router ReportPack The Daily Response Time line graph shows the average response time for the selected router for each day of the baseline period Daily Hourly Daily Response Time houston Acme s Support Office Fri Sep 10 1999 Thu Oct 21 1999 685 50 648 40 411 30 milliseconds 274 20 137 10 o Sep 10 Sep17 Sep24 Oct0O4 Oct O08 Oct 15 The Hourly Response Time line graph shows the response time for each sample for the report day for the selected router Daily Hourly Hourly Response Time houston Acme s Support Office Thu Oct 21 1999 1 216 00 972 00 729 00 486 00 milliseconds 243 00 o 00 00 03 00 06 00 09 00 12 00 15 00 18 00 21 00 00 00 6 20 Report Descriptions ReportPack Guide Data Source Source data for the Router ReportPack is collected using SNMP from several MIB II tables ifEntry system and ipAddrEntry Table 6 1 lists the variables that are polled from the MIB II ifEntry table Table 6 1 Variables Polled from MIB II ifEntry Table Variable OID String ifIndex A unique value for each interface Its value ranges between 1 and the value of ifNumber The value for each interface
125. ation from the daily exception table The Forecast report comprises the following components Report Descriptions 5 3 5 LAN and WAN ReportPacks The Estimated Days to Threshold DTT summary table ranks interfaces in descending order by 90 day projected utilization Estimated Days To Threshold DTT LAN Interfaces Exceeding or Within 90 Days Of Thresholds All Values Based On 95th Percentile Thresholds Util 40 Errors 5 Discards 5 Device Interface Description DTT Utilization Current P95 Projected DTT Percent Current P95 Utilization Util 90 Days Errors Percent Errors sanluis 2 Corporate Offices 52 00 34 37 43 83 104 00 6 07 miami 1 Sales 53 00 34 63 43 82 114 00 6 03 miami 2 hiktgs 54 00 34 41 43 57 92 00 6 09 sanjose 2 Mktgs 105 00 30 88 38 76 88 00 6 12 LO Double click on an interface to update the content of the following drill down components with performance data for the selected interface The Daily Grade of Service stacked bar chart shows GOS scores for the LAN metrics Utilization Errors and Discards and WAN metrics Input Utilization Output Utilization Total Percent Errors and Total Percent Discards over the baseline period for the selected interface For more information about the GOS weighting calculations for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively Daily Grade of Service sanjose Sales Tue Nov 09 1999 Mon Dec 20 1999 Critical Warning Good E
126. atus in several performance areas Drill down charts show daily distribution of bandwidth as polled volume against the baseline and congestion at the as polled level The following sections describe these charts in more detail For detailed information about the Frame Relay ReportPack see Frame Relay ReportPack on page 4 1 The QuickView and Snapshot reports comprise the following charts QuickView Daily Grade of Service chart see Daily Grade of Service QuickView on page 3 22 Snapshot Snapshot Selection table see Snapshot Selection Table Snapshot on page 3 23 Snapshot The Grade of Service chart see Grade of Service Snapshot on page 3 23 The Hourly Volume graph see Hourly Volume on page 3 24 The Hourly Congestion graph see Hourly Congestion on page 3 24 The Distribution of Bandwidth Utilization Input and Distribution of Bandwidth Utilization Output charts see Distribution of Bandwidth Utilization Input on page 3 25 and Distribution of Bandwidth Utilization Output on page 3 26 swioday wo gt lt S gt Ko J m z oO QuickView and Snapshot Reports 3 21 3 Analyzing TREND Reports Daily Grade of Service QuickView Figure 3 19 shows the top ten PVCs with the highest daily GOS scores Table 4 3 describes how Grade of Service is calculated Daily Grade of Service Frame Relay Interfaces with the Higest Grade of
127. average size of the transaction log in kilobytes over each day 1 of 2 Data Source 11 7 E m oz T ae p 2 oO zo is o p e 11 TREND Database ReportPack Table 11 1 TREND Database Daily Average and Maximum Variables Variable Description MAXlogsizekb The maximum size of the transaction log in kilobytes over each day AVGpct_logused The average percentage of the transaction log that is used over each day MAXpct_logused The maximum percentage of the transaction log that is used over each day AVGkb_logused The average number of kilobytes the transaction log used over each day MAXkb_logused The maximum number of kilobytes the transaction log used over each day 2 of 2 SD42SDdsi_database_util forecast Table For the SDdsi_database_util table TRENDsum calculates the average which is the baseline and the 30 60 90 day forecast for database used and transaction log used It also calculates the days to reach the 80 threshold for the percentage of da tabase used and percentage of the transaction log used Table 11 2 lists the variables Table 11 2 TREND Database Baseline and Forecast Variables Variable Description AVGAVGkb_dbused The average of daily averages for the number of kilo bytes used by the database over the baseline period 1 of 3 11 8 Data Source ReportPack G
128. ay ReportPack daily availability is calculated see Daily Availability Drill Down on page 3 42 Daily Hourly Daily Availability Sanjose Support Thu Now 11 1999 Wed Dec 22 1999 100 00 8 50 00 a a 0 Nov 11 Nov 18 Now25 Dec02 DecO9 Dec 16 Availability Service Level The Hourly Availability line graph compares the availability of the selected PVC with the Service Level Agreement over the baseline period Daily Hourly Hourly Availability Sanjose Support Wed Dec 22 12 00 AM Thu Dec 23 12 00 AM 100 20 g 66 80 g 2 a a 33 40 0 00 00 06 00 12 00 18 00 00 00 Availability Service Level 4 22 Report Descriptions ReportPack Guide The PVCs with Highest Response Time for the Day bar chart shows the PVCs with the highest average response times for the previous day For more information about how network response times are calculated see Response Time on page 3 45 Response Time PVCs With Highest Response Time Wed Dec 22 1999 Device Interface DCLI Resp Time COPBKR_BCIO1 202104 0 23 216 00 COPBKR_BCIO1 202104 0 17 216 00 COPBKR_BCIO1 202104 0 24 215 00 COPBKR_BCIO1 202104 0 16 215 00 COPBKR_BCIO1 202104 0 19 214 00 COPBKR_BCIO1 202104 0 25 214 00 COPBKR_BCIO1 202104 0 26 213 00 COPBKR_BCIO1 202104 0 22 212 00 COPBKR_BCIO1 202104 0 28 212 00 Double click on a device to update the content of the following drill down components with performance data for t
129. ber of days in the rolling baseline period is reached processing continues in the fashion described above However the rolling baseline value that is computed from the preceding 42 days of samples is appended to the table The existing baseline is no longer replaced Thus rows are added to the rolling baseline summary table as follows Day Baseline Statistic Computed from 1 ay Days thru 42 2 az Days 2 thru 43 3 a3 Days 3 thru 44 4 a4 Days 4 thru 45 n an Days n thru n 41 The Executive Summary reports in the TREND ReportPacks compare a resource usage for each hour of the day to an hourly baseline that has been computed for the resource The concept for computing an hourly baseline is the same as the one for computing a daily baseline However the hourly baseline table has one row for each hour of each day computed for a sliding 42 day baseline period as described above Baselines ReportPack Guide Ranking Ranking lets you chart the volatility of a given metric for a given element For example suppose you want to use utilization and congestion among 100 PVCs to answer the following questions Which ten PVCs rank the highest in terms of utilization and congestion on any given day How often does the same element appear in the top ten in terms of utilization and congestion Are there significant variations in element rank from day to day delta rank N ie lt
130. by most routers Channel Service Units CSUs yoed 4 1 A TI o 3 D D o o lt D o fe T 4 Frame Relay ReportPack Data Service Units DSUs and other network devices For more information about the Frame Relay ReportPack data source see Data Source on page 4 25 Report Descriptions The following sections describe each report in the Frame Relay ReportPack in detail Executive Summary Report The Executive Summary report provides executives and high level managers with graphs that aggregate key metrics for all elements within a class They can review the daily volume summary and if desired examine more detailed information about volume grade of service utilization congestion and total exception count Note The TRENDweb Administrator can configure a Frame Relay ReportPack installation so that the Executive Summary report will only display data for devices belonging to a specific customer For more information about this feature see Displaying Data for Specific Customer Devices in Reports on page 1 20 The Executive Summary report comprises the following charts The Daily Volume for PVCs bar chart shows the total Frame Relay volume over the baseline period by default the six weeks prior to and including the 4 2 Report Descriptions ReportPack Guide report date giving the executive an instant view of increases or decrease in Frame Relay usage over time PYC Mana
131. centage of the transaction log used for the baseline period and the projected percent usage of the transaction log for the 30 60 and 90 day forecasts Transaction Log Max Utilization 30 60 and 90 Day Forecast Baseline Tue Jun 15 1999 Mon Jul 26 1999 Percent Utilized 00 00 00 H baseline ULTRA_SYBASE 0 30 Day ULTRA_SYBASE 0 60 Day ULTRA_SYBASE 0 90 Day ULTRA_SYBASE 0 The Database Usage and Size line graph shows the number of kilobytes for the daily average and maximum database usage for the last 90 days It also shows the database size in kilobytes Database Usage and Size Average and Maximum Last 90 Days Wed Apr 28 1999 Mon Jul 26 1999 4 069 976 00 802 482 00 534 988 00 Kilobytes 267 494 00 0 Apr 28 May12 May26 Jun09 Jun23 JulO Jul 21 Daily Average Usage ULTRA_SYBASE 0 Daily Maximum Usage ULTRA_SY BASE 0 Database Size ULTRA_SYBASE 0 Report Descriptions ReportPack Guide The Database Average Usage bar chart shows the average number of kilobytes the database used for the baseline period and the number of kilobytes the database will use for the 30 60 and 90 day forecasts Database Average Usage 30 60 and 90 Day Forecast Baseliine Tue Jun 15 1999 Mon Jul 26 1999 756 441 60 504 294 40 Kilobytes 252 147 20 00 00 00 Wi baseline ULTRA_SYBASE 0 30 Day ULTRA SYBASE 0 m60 Day ULTRA SYBASE 0 90 Day ULTRA_SYBASE 0 The Transaction Log U
132. centage of time over the sample period the system s CPU was executing in user mode cpuStatsSysPercent 1 3 6 1 4 1 546 13 1 1 10 2 of 3 12 44 Data Source ReportPack Guide Table 12 3 System Variables Polled from Empire cpuStatsEntry Table Variable OID String executing the kernel or operating system The percentage of time over the sample period that the system s CPU was cpuStats WaitPercent 1 3 6 1 4 1 546 13 1 1 11 operating system The percentage of time over the sample period the system s CPU was waiting for I O This object may not be supported by the underlying 3 of 3 Table 12 4 lists the variables from the diskStatsEntry table in the empire mib file Table 12 4 System Variables Polled from Empire diskStatsEntry Table Variable OID String diskStatsIndex 1 3 6 1 4 1 546 12 1 1 1 An index that uniquely identifies an entry in the diskStatsTable diskStatsQueueLength 1 3 6 1 4 1 546 12 1 1 2 the last measurement period The average number of operations waiting in the disk s service queue over diskStatsServiceTime 1 3 6 1 4 1 546 12 1 1 3 1 of 3 Data Source 12 45 _ N w lt P 3 D O U o Q 12 System ReportPack 12 46 Table 12 4 System Variables Polled from Empire diskStatsEntry Table Variable OID String The average ser
133. centile of the value with the Grade of Service weight applied for the selected Cisco router for each day of the baseline period Report Descriptions ReportPack Guide The Daily CPU Utilization line graph shows the weighted average and 95th percentile CPU utilization for the selected Cisco router for each day of the baseline period Daily CPU Utilization tulsa Desktalk Support 4 hon Now 08 1999 Sun Dec 19 1999 123 20 92 40 61 60 Percent 30 80 Nov 22 Dec 06 P95 CPU Utilization Awg CPU Utilization The Daily Memory Utilization line graph shows the weighted average and 95th percentile memory utilization for the selected Cisco router for each day of the baseline period Memory Buffer Daily Memory Utilization tulsa Desktalk Support 4 Mon Nov 08 1999 Sun Dec 19 1999 108 00 72 00 36 00 Nov 22 Dec 06 P95 Memory Utilization Awg Memory Utilization Report Descriptions 8 21 D o S e U o i x J9INOY ODSID 8 8 Cisco Router ReportPack The Daily Buffer Utilization graph shows the weighted average of the weighted average of the buffer utilization samples collected for the selected Cisco router for each day of the baseline period for each buffer size There are four buffer sizes huge large medium and small Memory Buffer Daily Buffer Utilization tulsa Desktalk Support 4 hon Now 08 1999 Sun Dec 19 1999 Nov 22 Dec 06 Huge Buffer Utilization
134. ces Chapter 8 Cisco Router ReportPack focuses on the Cisco router interfaces Chapter 9 RMON ReportPack focuses on the various segments of Ethernet networks Token Ring networks or both Chapter 10 RMON2 ReportPack focuses on data collected from RMON2 probes Chapter 11 TREND Database ReportPack focuses on the historical and predicted usage levels of the TREND database and transaction log Chapter 12 System ReportPack focuses on statistics about CPU utilization paging and swapping file system usage virtual and physical memory cache hits and misses and grade of service specifically for the Empire SystemEDGE agent Appendix A describes the property file formats for various ReportPacks ReportPack Guide Related Documents This guide is designed to help you understand each of the ReportPacks and use them to accurately analyze the data they contain Use this guide in conjunction with the following guides TREND User s Guide TREND Installation Guide Guide to Viewing TREND Reports Guide to Building TREND Reports Online User Manual Availability TREND is shipped with online viewable user manuals in Portable Document Format PDF along with the Adobe Acrobat Reader These documents and the Adobe Acrobat Reader are available prior to installation in the Documents directory of the installation CD You need to install the Adobe Acrobat Reader to view these documents online xvii Preface
135. change report also may highlight which newly installed devices and technologies are approaching a logical threshold and which may be exceeding a threshold or which are underutilized Drastic increases in grade of service scores suggest additional PVCs to be watched while decreases document an effective remedy to a previously negative condition The Top Ten Report comprises the following components o gt D fe c D D me fe U o P x The Routers with Highest Throughput table sorts by rank up to ten routers by the previous day s throughput in ascending order Routers With Highest Throughput Thu Oct 21 1999 Router Pkts sec Rank Bytes sec Pct Discards boston 37741747 27 1 1213756965 47 1 86 tulsa 35379888 04 2 947004189 12 2 48 lincoln 33926123 71 3 443368212 95 3 73 sanjose 27538522 93 4 1182195575 79 3 73 miami 27480781 14 5 1182570806 32 3 73 sanluis 27366425 47 6 8534497 19 02 3 48 houston 11812231 66 7 972846001 10 2 58 cotati 4015947 52 8 174640564 83 2 52 The Routers with Greatest Change in Throughput table sorts by change in rank up to ten routers by previous day s throughput change in ascending order This table may highlight which newly installed devices and technologies are approaching a logical threshold may be exceeding a threshold or may be impeding optimum utilization Report Descriptions 6 15 6 Router ReportPack 6 16 Router Sanjose tulsa miami boston houston li
136. chitecture of the property import feature D U D 2 gt 2 3 2 D 1 Figure 1 3 Architecture of the Property Import Feature Step 1 Data is exported from the key tables into tab delimited files The tab delimited files are edited and imported into the property data tables Then an update procedure runs to populate the property columns in the key tables If you already have a tab delimited skip Step 1 and copy your files into the PropertyData directory The lt ReportPackName gt _exportdata pro file runs to export the data Data is exported from the key table s into PropertyData directory The lt RepatPackvame gt _importdata pro file runs to import the data and run the update script Step 3 NS SHOR 2 Data from the ReportPack s Data is edited and then property data tables are used to update the property values in the key tables imported into the ReportPack s property data tables Data is exported from the key tables to a property file which is a tab delimited ASCII file You can generate it using the lt Report Pack gt _exportdata pro process or you can create it using an application like Microsoft Excel Then you edit the data and import the file into the ReportPack s property data tables The data is used to update property information contained in the key tables The format of the the property file is uniquely de
137. ck 3 27 4 9 for LAN ReportPack 5 13 5 15 for System ReportPack 12 14 for WAN ReportPack 5 13 5 15 P Property Import feature using 1 14 1 24 property import feature property file formats A 1 A 5 for A 3 for Bay Router A 4 for Cisco Router A 4 for LAN WAN A 2 for Router A 4 Q QuickView report description of 2 5 for Bay Router ReportPack 7 9 for Cisco Router ReportPack 8 10 for Frame Relay ReportPack 3 20 4 7 for LAN ReportPack 5 6 for RMON ReportPack 9 3 for RMON ReportPack 10 5 for Router ReportPack 6 8 for System ReportPack 12 9 for WAN ReportPack 5 6 R Ranking definition of 2 13 delta rank definition of 2 13 ranking table 2 13 removing ReportPacks 1 25 1 26 ReportPacks administration of 1 1 1 28 dependencies of 1 27 directory structure of 1 4 installing 1 2 1 14 removing 1 25 1 26 unzipping sources for 1 3 reports analyzing 3 1 3 47 Capacity Planning 3 13 3 17 Executive Summary 3 2 3 7 Forecast 3 7 3 12 Hot Spots 3 32 3 40 Service Level Management 3 41 3 47 Top Ten 3 17 3 20 components of 2 3 data displaying by customer 1 20 importing properties 1 14 metric classes for 2 6 snapshot selecting targets elements for 1 27 standard types of 2 4 Capacity Planning 2 5 Executive Summary 2 4 Forecast 2 4 Hot Spots 2 4 QuickView 2 5 ReportPack Guide Service Level Management 2 5 Snapshot 2 5 Top Ten 2 5 RMON ReportPack description of 9 1 installing 1 2 report description
138. d all ones broadcast address is used the value will be 1 This value applies to both the subnet and network broadcasts addresses used by the entity on this logical ipAdEntReasmMaxSize 1 3 6 1 2 1 4 20 1 5 entreasmmaxsize013 The size of the largest IP datagram which this entity can reassemble from incoming IP fragmented datagrams received on this interface 6 26 Data Source ReportPack Guide Basic Metric Calculations The performance metrics used in Router reports are based on the calculations listed in Table 6 4 Note that some charts modify these calculations Table 6 4 Router Metrics Metric Computed As Discards discard_pct ifindiscards021 ifoutdiscards027 ifinucastpkts019 ifinnucastpkts020 ifindiscards02 1 ifinerrors022 inunknownprotos023 ifoutnucastpkts026 ifoutucastpkts025 1 100 0 The ratio of the sum of inbound and outbound packets containing errors and those that are discarded because they come from an unknown source to the total packet traffic inbound and outbound at the interface plus 1 In this computation a negligible amount that is 1 packet is added to the divisor to prevent a potential divide by zero condition The result is multiplied by 100 to yield a percentage o gt D fe c D me e V o ie x Throughput TOT ifinucastpkts ifoutucastpkts delta_time The total number of ifinucastpkts plus ifoutucastpkts
139. d not be used to judge the amount of the system load Managers should also monitor swapCapacity 9 of 9 The MIB II server_volume table is a miblet of the MIB II ifEntry table Table 12 2 lists the variables Table 12 2 System Variables Polled from MIB II Server Volume Table Variable OID String ifSpeed 1 3 6 1 2 1 2 2 1 5 An estimate of the interface s current bandwidth in bits per second For interfaces that do not vary in bandwidth or for those where no accurate estimation can be made this object should contain the nominal bandwidth ifInOctets 1 3 6 1 2 1 2 2 1 10 The total number of octets received on the interface including framing characters ifOutOctets 1 3 6 1 2 1 2 2 1 16 1 of 2 12 42 Data Source ReportPack Guide Table 12 2 System Variables Polled from MIB II Server Volume Table Variable OID String The total number of octets transmitted out of the interface including framing characters 2 of 2 Table 12 3 lists the variables from the cpuStatsEntry table in the empire mib file Table 12 3 System Variables Polled from Empire cpuStatsEntry Table Variable OID String cpuStatsIndex 1 3 6 1 4 1 546 13 1 1 1 An index that uniquely identifies an entry in the cpuStatsTable cpuStatsDescr 1 3 6 1 4 1 546 13 1 1 2 The type of CPU for which statistics are reported On Solaris 2 x this description also incl
140. ded packets ifindiscards ifoutdiscards for the selected Bay router over each hour of the report day Total Number of Discarded Packets sanjose Sun Dec 19 1999 3 754 805 00 3 003 844 00 2 252 883 00 1 501 922 00 Total Discards 750 961 00 0 00 00 06 00 12 00 18 00 00 00 7 12 Report Descriptions ReportPack Guide Near Real Time QuickView and Snapshot Reports The Near Real Time Quick View report provides router performance statistics up to the last SNMP poll The report does not rely on nightly summaries therefore it provides instant reporting on collected data Network managers and analysts can go to this report to select a router from a table and assess its status in the following performance areas CPU utilization buffer utilization and memory utilization Alternatively the Snapshot report shows the same information however you select routers you want to view from a pick list when you invoke the report For more information about selecting a router for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Note For an in depth analysis of how to interpret the Bay Router Near Real Time QuickView and Snapshot reports and for chart illustrations see Near Real Time Quick View and Snapshot Reports on page 3 27 The QuickView and Snapshot reports comprise the following charts The Router Selection List table shows each router and its performance sta
141. desired look into more detailed information such as grade of service hourly volume compared to baseline hourly grade of service and total exception count The Executive Summary report comprises the following components The Daily Network Volume bar chart shows the total volume for all systems over each day of the baseline period Daily Network Volume All Systems Mion Jan 17 2000 Mon Feb 28 2000 3 082 50 2 466 00 1 849 50 Terabytes 1 233 00 616 50 0 Sep 10 Sep14 Sep18 Sep22 Sep26 Sep30 OctO4 Otos Oct 12 Oct 16 Oct 20 12 2 Report Descriptions ReportPack Guide The Hourly Network Volume vs Baseline bar chart shows the average volume for all systems for each hour of the report day compared to the baseline volume of all systems Hourly Network Volume vs Baseline All Systems Thu Oct 21 1999 176 997 20 132 747 90 88 498 60 Gigabytes 44 249 30 0 00 00 06 00 12 00 18 00 Blume in Gigabytes Baseline volume Al Devices 0 The Total Exception Counts stacked bar chart shows the number of times a system exceeded a queue length or CPU memory or swap utilization threshold during each hour of the report day Total Exception Counts All Systems Thu Oct 21 1999 81 00 54 00 Total 27 00 it 00 00 06 00 12 00 18 00 E Queue Length B swap Dtemory Utilization E CPU Utilization The Hourly Grade of Service stacked bar chart provides an overview of system performance b
142. discards for each hour during the report day For more information about how Discards is computed see Table 7 2 Report Descriptions 8 13 D o S e U o i x J9 NOY ODSID 8 8 Cisco Router ReportPack Percent Discards boston Acme s Personnel Office Sun Dee 19 1999 2 40 1 80 1 20 Percent 0 60 0 00 00 06 00 12 00 18 00 00 00 Snapshot Report For more information about selecting a router for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The Snapshot report comprises the following components The Snapshot Selection table lists the time period for which the router information was collected the router name any associated interfaces and the router s CPU Utilization Memory Utilization and Buffer Utilization Double click on a router in the table to update the content of the following components with performance data for the selected router Snapshot Selection Sun Dec 19 1999 Router Interface Description CPU Util 95th Memory Util 95th Buffer Util 95th percentile percentile percentile ae ee ee cotati 0 Acme s Sales Office 7 00 98 07 93 78 houston 0 Acme s Support Office 98 00 97 98 22 90 lincoln 0 Desktalk Support 1 46 00 87 56 59 02 tulsa o Desktalk Support 4 94 00 80 62 44 87 8 14 Report Descriptions ReportPack Guide The Hourly Grade of Service bar chart shows the Grade of Service scores for Physical Memory U
143. e in hundredths of a second s of the system was last reinitialized ince the network management portion sysContact 1 3 6 1 2 1 1 4 The textual identification of the conta together with information on how to contact this person ct person for this managed node sysName 1 3 6 1 2 1 1 5 An administratively assigned name for this managed node By convention this is the node s fully qualified doma in name sysLocation 1 3 6 1 2 1 1 6 The physical location of this node for example telephone closet 3rd floor sysServices 1 3 6 1 2 1 1 7 2 of 3 Data Source ReportPack Guide Table 6 2 Variables Polled from MIB II System Table Variable OID String A value which indicates the set of services that this entity primarily offers The value is a sum This sum initially takes the value zero Then for each layer L in the range 1 through 7 that this node performs transactions for 2 raised to L 1 is added to the sum For example a node which performs primarily routing functions would have a value of 4 24 3 1 In contrast a node which is a host offering application services would have a value of 72 24 4 1 24 7 1 Note that in the context of the Internet suite of protocols values should be calculated accordingly 1 layer functionality 2 physical for example repeaters 3 datalink subnetwork for example bridges 4 internet for
144. e an overloaded system numPageIns 1 3 6 1 4 1 546 1 1 7 8 22 The total number of pages that have been paged in since the kernel was last initialized Page in is an operation performed by the virtual memory system in which the contents of a page are read from secondary storage A high rate of change may indicate an overloaded system numPageOuts 1 3 6 1 4 1 546 1 1 7 8 23 The total number of pages that have been paged out since the kernel was last initialized Page out is an operation performed by the virtual memory system in which the contents of a page are written to secondary storage A high rate of change may indicate an overloaded system numPageReclaims 1 3 6 1 4 1 546 1 1 7 8 24 6 of 9 P i lt a o 3 D xe e Data Source 12 39 5 S 2 12 System ReportPack Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The total pages that have been reclaimed from the freelist since the kernel was last initialized Memory pages are stored on a free list when not in use If a page has been acquired from the free list it has been reclaimed A high rate of changed indicates an active system which is not generally consid ered to be overloaded numPageFaults 1 3 6 1 4 1 546 1 1 7 8 25 The total number of page faults that have occurred since the kernel was last initialized A page fault occurs when a running process attempts to access a
145. e previous day A large change in rank identifies a volatile resource and you should investigate the reasons for the high change in discard percentage Bay Routers With Greatest Change in Percentage of Discards Wed Jan 05 2000 Device Pct Pkts Pkts sec Rank Bytes sec Change Rank Previous Pct Prior Rank Discarded Pkts Discarded sanluis 3 48 27366425 47 4 8534497 19 02 4 3 48 4 Sanjose 3 73 27538522 93 3 1182195575 79 6 3 73 Figure 3 18 Top Ten Bay Routers with the Greatest Change in Percentage of Discards QuickView and Snapshot Reports 3 20 The Quick View report identifies the ten elements with the highest worst grade for scored metrics Network managers and analysts can go to this report to select an element and assess its status in several performance areas Alternatively the Snapshot report shows the same information however you select elements you want to view from a pick list when you invoke the report For more information about selecting an element for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 For example the Quick View and Snapshot reports for the Frame Relay ReportPack identify the top ten PVCs with the highest worst grade of service scores according to Grade of Service score Network managers and analysts can go to this report to QuickView and Snapshot Reports ReportPack Guide select a PVC from a grade of service chart and assess its st
146. eak percent collisions for the past 12 months if available for the interface selected in the Monthly RMON Ethernet Statistics table RMON Ethernet Collisions bluegrass 1_ALL Pug 1999 Aug 1999 o D fe Z pej fe U o Q Percent Collisions 20 00 00 E Avg Pet Collisions M Peak Pct Collisions Daily RMON Token Ring The Daily RMON Token Ring comprises the following components The Daily RMON Token Ring Statistics table provides a list of all Token Ring interfaces with their daily average and peak utilization average beacons Report Descriptions 9 7 9 RMON ReportPack per second and total beacons The interfaces are sorted in descending order based on daily peak utilization Daily RMON Token Ring Statistics Sorted by Peak Utilization Wed Oct 13 1999 Wed Oct 13 1999 RMON Probe LF Speed Avg Pct Util Peak Pct Util Avg Beacons Total Beac Sec bluegrass 110_BI_ALL 64000 64000 0 bluegrass 21_ALL 64000 64000 0 0 30 1 32 0 00 0 00 bluegrass 110_DTE_ALL 64000 64000 0 0 30 1 32 0 00 0 00 bluegrass 22_ALL 64000 64000 0 0 26 0 86 0 00 0 00 bluegrass 110_DCE_ALL 64000 64000 0 0 26 0 85 0 00 0 00 bluegrass 41_ALL 10000000 1 0E7 0 04 0 62 0 03 2509 6 tms east desktalk com 41_ALL 10000000 1 0E7 0 02 0 25 0 01 457 33 Double click on an interface in the table to update the content of the following components with performance data for the selected interface The RMON Token Ring
147. ecast U boston 12 00 99 02 105 58 112 14 128 00 81 62 82 87 lincoln 1000 00 47 041 47 60 48 20 545 00 88 96 89 69 houston 0 00 98 00 98 00 98 00 1000 00 98 15 98 26 cotati N A 7 02 7 01 6 99 1000 00 98 10 98 18 LO The Underutilized Cisco Routers table ranks the Cisco routers projected to be the least utilized in terms of CPU Utilization 10 Memory Utilization 10 or Buffer Utilization 10 within the next 90 days in ascending order Underutilized Cisco Routers Projected To Be Least Utilized For 90 Days Thresholds CPU 10 Phys Mem 10 Buffer 10 Router DTT CPU 30 Day CPU 60 Day CPU 90 Day CPU DTT Memory Util 30DayMem 60DayMem 9 Utilization Util Forecast Util Forecast Util Forecast Util Forecast Util Forecast U p_e Double click on a router in either table to update the content of the following components with performance data for the selected router Report Descriptions D o S e U o i x J9 NOY ODSID 8 8 Cisco Router ReportPack 8 20 The Daily GOS chart shows grade of service scores for CPU Utilization Memory Utilization Buffer Utilization and Buffer Errors for the selected Cisco router for each day of the baseline period Daily GOS tulsa Desktalk Support 4 hon Nov 08 1999 Sun Dec 19 1999 Critical Warning Good Excellent Now 08 Nov 22 Dec 06 M buffer Faults E Buffer utilization Hnemory Utilization W CPU Utilization Each metric shows the 95th per
148. ect contains a zero value ifInOctets 1 3 6 1 2 1 2 2 1 10 The total number of octets received on the interface including framing characters ifInUcastPkts 1 3 6 1 2 1 2 2 1 11 The number of subnetwork unicast packets delivered to a higher layer protocol ifInNUcastPkts 1 3 6 1 2 1 2 2 1 12 The number of non unicast i e subnetwork broadcast or subnetwork multicast packets delivered to a higher layer protocol ifInDiscards 1 3 6 1 2 1 2 2 1 13 The number of inbound packets which were chosen to be discarded even though no errors had been detected to prevent their being deliverable to a higher layer protocol One possible reason for discarding such a packet could be to free up buffer space ifInErrors 1 3 6 1 2 1 2 2 1 14 3 of 5 Data Source 5 29 5 LAN and WAN ReportPacks Table 5 1 LAN and WAN Connectivity Variables Variable OID String The number of inbound packets that contained errors preventing them from being deliverable to a higher layer protocol iffnUnknownProtos 1 3 6 1 2 1 2 2 1 15 The number of packets received via the interface which were discarded because of an unknown or unsupported protocol ifOutOctets 1 3 6 1 2 1 2 2 1 16 The total number of octets transmitted out of the interface including framing characters ifOutUcastPkts 1 3 6 1 2 1 2 2 1 17 The total number of packets that higher level protocols
149. ect routers you want to view from a pick list when you invoke the report For more information about selecting a router for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The QuickView and Snapshot reports comprise the following charts Report Descriptions 8 15 D D xe e Uv o i x JO INOY ODSID 8 8 Cisco Router ReportPack The Router Selection List table shows each router and its performance statistics based on the data polled for the last 6 hours Depending on the type of statistic the value shown is either the average or maximum value of all the data collected over the 6 hour period Router Selection List Select interface to see near real time reports Router Avg CPU Max CPU Small Buffer Medium Buffer Big Buffer Large Buffer Huge Buffer Utilization Utilization Utilization Utilization Utilization Utilization Utilization Sa a ee enn tulsa 39 70 68 00 56 25 50 00 50 00 50 00 76 75 boston 38 25 67 00 56 25 50 00 50 00 50 00 76 75 atlanta 34 30 58 00 75 00 75 00 73 33 68 75 86 67 lincoln 19 70 35 00 12 50 60 00 50 00 50 00 30 00 cotati 2 80 5 00 6 25 11 00 5 00 4 17 8 33 Double click on a device in the table to update the content of the following drill down components with performance data for the device The Average CPU Utilization line graph shows the average CPU utiliza tion for each sample taken during the report period for the selected router
150. edure file and run it using the following command trend_proc f lt ReportPackName gt _exportdata pro This process calls trend_export which reads a definition file to determine which columns to export from the database It exports a template property file The file contains a row for each object that is currently discovered by the ReportPack If no objects are discovered then the file contains no rows Therefore this utility is only useful after discovery and at least one polling cycle has been completed All exported columns come from the key table s Trend_export generates a tab delimited property file and place it in the following folder DPIPE_HOME packages lt ReportPackName gt lt ReportPackName gt ap PropertyData After you generate the file you can edit it manually or using an application like Microsoft Excel and then export the data to a tabbed delimited file Creating Your Own Property File It does not matter what application you use to create the property file However the property file must be an ASCII tab delimited file and the format of the file must comply with one of the column specifications in Property File Formats on page A 1 Using the Property Import Feature ReportPack Guide An example property file is provided for each ReportPack It is located in the following directory DPIPE_HOME packages lt ReportPack gt lt ReportPack gt ap PropertyData D
151. egative condition Report Descriptions ReportPack Guide LAN Interfaces With Greatest Change In Grade Of Service GOS Scores Mon Dec 20 1999 Device Interface Description GOS Score Previous GOS Change Rank GOSRank Previous Rank U GOS Score houston 2 Accounting 1 52 1 54 1 17 18 houston 3 Stock Room 1 99 1 97 2 7 10 tulsa 2 Tulsa Research Center 1 54 1 53 4 16 19 miami 1 Sales 1 96 1 97 8 10 11 sanjose 2 Niktgs 1 87 1 86 11 13 15 sanjose 1 Sales 1 95 1 95 12 41 13 miami 2 Mktgs 1 95 1 95 13 12 14 boston 2 Northeast Sales 1 98 1 98 14 8 9 sanluis 2 Corporate Offices 1 96 1 96 16 9 12 cotati 2 Cotati Sales 1 00 1 00 19 18 20 LO Service Level Management Report The Service Level Management report lets CIOs CFOs network managers and customers know if their interfaces are meeting contracted service levels for availability and response times by reporting interface availability across the enterprise over time and for individual interfaces for the day It reports the ten interfaces with lowest availability for the day and the ten interfaces with the highest network response times It also provides drill down reports for daily availability over the baseline period and daily and hourly response times for the selected interfaces with lowest availability or highest network response time The Service Level Management report comprises the following components The LAN Interfaces with Lowest Availability for the Day ba
152. elated tools which can provide detailed information about the health and performance of a network TREND helps a network administrator troubleshoot problem areas maximize performance and conduct historical and trending analysis The TREND reports that are described in this guide are the key to successful network administration This guide can help you accurately analyze the data they contain For information about generating these reports see the TREND User s Guide Organization of This Guide The chapters of this guide are organized as follows Chapter 1 describes installation of ReportPacks dependencies between ReportPacks and selection of targets elements for Snapshot reports Chapter 2 describes the uniform layout of the reports and lists the eight reports common to most ReportPacks XV Preface xvi Chapters 3 analyzes selected reports from each ReportPack to give you guidelines for interpreting the information provided in any one report type Chapters 4 12 describe each ReportPack in detail including descriptions of each report each chart within a report data sources supported devices and metric and GOS calculations Chapter 4 Frame Relay Report Pack focuses on the Frame Relay switch and PVCs Chapter 5 LAN and WAN ReportPacks focuses on LAN and WAN devices Chapter 6 Router ReportPack focuses on the various router interfaces Chapter 7 Bay Router ReportPack focuses on the Bay router interfa
153. eline period for each buffer size There are four buffer sizes huge large medium and small Memory Buffer Daily Buffer Utilization tulsa Desktalk Support 4 tion Nov 08 1999 Sun Dec 19 1999 Percent N a p ta Nov 08 Nov 22 Dec 06 Huge Buffer Utilization Large Buffer Utilization Medium Buffer Utilization Small Buffer Utilization The Daily Total Volume bar chart shows the total number of bytes volume for the selected Cisco router for each day of the baseline period For more information about how Volume is computed see Table 7 2 Total Bytes Discards Daily Total Volume tulsa Desktalk Support 4 Mion Now 08 1999 Sun Dec 19 1999 57 917 577 20 43 438 182 90 28 958 783 60 Megabytes 14 479 394 30 0 Nov 08 Nov 22 Dec 06 Report Descriptions 8 9 D D S e U o i x JO INOY ODSID 8 8 Cisco Router ReportPack The Daily Percent Discards bar chart shows the percentage of discards for the selected Cisco router for each day of the baseline period For more information about how Discards is computed see Table 7 2 Total Bytes Discards Daily Percent Discards tulsa Desktalk Support 4 hon Now 08 1999 Sun Dec 19 1999 3 00 Percent 0 Nov 08 Nov 22 Dec 06 QuickView and Snapshot Reports The Quick View report provides detailed information on individual Cisco Routers in a single report The Snapshot report provides the same information as
154. en Ring Statistics Sorted by Peak Utilization LF Speed Avg Pct Util 110_Bl_ALL 64000 64000 0 21_ALL 64000 64000 0 110_DTE_ALL 64000 64000 0 22_ALL 64000 64000 0 110_DCE_ALL 64000 64000 0 41_ALL 10000000 TOEF 1_ALL 10000000 1 0E7 1_ALL 10000000 1 0E7 Pug 1999 0 55 0 30 0 29 0 26 0 26 0 03 0 02 0 02 Peak Pct Util 1 86 1 35 1 35 1 00 1 00 0 62 0 25 0 24 Avg Beacons Total Beacons Sec 0 00 0 00 0 00 0 00 0 00 0 02 0 00 1 23 0 00 0 00 0 00 0 00 0 00 37020 65 10046 19 2894206 53 Double click on an interface in the table to update the content of the following components with performance data for the selected interface The RMON Token Ring Utilization bar chart shows the monthly average 95th percentile and peak utilization for the past 12 months if available for the interface selected in the Monthly RMON Token Ring Statistics table Percent Utilization Pug 1999 Aug 1999 20 00 00 E Avg Pet Util M Peak Pct Util RMON Token Ring Utilization bluegrass 3_ALL 64000 Report Descriptions ReportPack Guide Report Descriptions The RMON Token Ring Beacons bar chart shows the monthly average and peak beacons per second for the past 12 months if available for the interface selected in the Monthly Token Ring Ethernet Statistics table o Pe fe Z pej 3 fe U o Q RMON Token Ring Beacons bluegrass 3_ALL 64000 fug 1999 Aug 1999
155. en they require the attention of the system An excessive rate of interrupts can indicate a heavily loaded machine servers and machines shared by many users will typically have a higher rate though numPageSwapIns 1 3 6 1 4 1 546 1 1 7 8 18 The total number of pages that have been swapped in since the kernel was last initialized Page swapping occurs when a process is swapped to and from secondary storage A high rate of page swap activity can indicate an overloaded system numPageSwapOuts 1 3 6 1 4 1 546 1 1 7 8 19 The total number of pages that have been swapped out since the kernel was last initialized Page swapping occurs when a process is swapped to and from secondary storage A high rate of page swap activity can indicate an overloaded system numSwapIns 1 3 6 1 4 1 546 1 1 7 8 20 5 of 9 Data Source ReportPack Guide Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The total number of processes that have been swapped in since the kernel was last initialized Processes are swapped to and from secondary storage when memory is in short supply A high rate may indicate an overloaded system numSwapOuts 1 3 6 1 4 1 546 1 1 7 8 21 The total number of processes that have been swapped out since the kernel was last initialized Processes are swapped to and from secondary storage when memory is in short supply A high rate may indicat
156. er Hetwork Layer Rev Mbytes Sent Mbytes 134 70 32 5 bluegrass 3 ip 1 72 02 72 49 134 70 32 251 bluegrass 3 ip 1 52 70 4 47 134 70 150 17 bluegrass 3 ip 1 29 32 26 09 134 70 32 24 bluegrass 3 ip 1 17 43 25 44 209 219 153 173 bluegrass 3 ip 1 15 91 15 47 134 70 65 106 bluegrass 3 ip 1 15 85 16 01 134 70 180 100 bluegrass 3 ip 1 15 45 15 02 134 70 65 206 bluegrass 3 ip 1 15 26 14 89 134 70 180 113 bluegrass 3 ip 1 14 32 13 93 134 70 32 56 bluegrass 3 ip 1 12 35 0 82 Report Descriptions 10 11 10 RMON2 ReportPack Weekly Top Ten The Weekly Top Ten report lists the top ten conversation pairs talkers and listeners for the preceding week The Weekly Top Ten report comprises the following components The Top Conversation Pairs table lists the source and destination IP address pairs that sent the most data on the preceding week along with the total megabytes sent The Top Ten Talkers table lists the IP address and the total megabytes sent and received for the addresses that sent the most data to all destinations on the preceding week The Top Ten Listeners table lists the IP address and the total megabytes sent and received for the addresses that received the most data from all destinations on the preceding week 10 12 Report Descriptions 11 TREND TREND Database ReportPack The TREND Database ReportPack reports historical and predicted usage levels of the TREND database and transaction log It provides a too
157. er Inventory summary table lists the top ten routers with the highest throughput pkts sec in your enterprise Quick View or only for routers that you select Snapshot For more information about selecting targets or elements for a snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Router Inventory Thu Oct 21 1999 Router Vendor WAH Interfaces LAH Interfaces Other Total Pkts sec Interfaces Interfaces tulsa Cisco 2 1 o 3 35379888 04 lincoln Cisco 1 N A 0 4 33926123 71 sanjose Bay 4 2 o 3 27538522 93 miami Cisco 4 2 0 3 27480781 14 sanluis Bay 1 1 0 2 27366425 47 6 8 Report Descriptions ReportPack Guide The Hourly Volume bar chart shows the volume of data flow for each hour of the report day for the selected router Hourly Volume boston Acme s Personnel Office Thu Oct 21 1999 4 703 50 o gt D fe o oO ze fe o Thu 00 00 Thu 09 00 Thu 18 00 AS O 2 The Packets Per Second line graph shows the packets second over each hour in the report day for the selected router Packets Per Second boston Acme s Personnel Office Thu Oct 21 1999 6 437 796 00 150 236 80 3 862 677 60 2 575 118 40 Packets Second 1 287 559 20 0 00 00 06 00 12 00 18 00 00 00 Report Descriptions 6 9 6 Router ReportPack The Discards bar chart shows the number of discards over each hour in the day for the selected router Disca
158. erefore they provide instant reporting on collected data The QuickView lets you select a device from a table and assess its status in several performance areas wo gt lt S J Ko J m z oO Alternatively the Snapshot report shows the same information as the Quick View report however you select the device you want to view from a pick list when you invoke the Snapshot report For more information about selecting a device for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Near Real Time QuickView and Snapshot Reports 3 27 3 Analyzing TREND Reports The following ReportPacks provide Near Real Time QuickView and Snapshot reports Bay Router Cisco Router Frame Relay LAN WAN gt System For more information about the Near Real Time reports for each ReportPack see the appropriate chapter for the ReportPack in this guide For example the Near Real Time QuickView and Near Real Time Snapshot reports for the Bay Router ReportPack comprises the following table and graphs Router Selection List table see Router Selection List on page 3 29 Double click on a router in the table to update one of the following drill down charts with performance data for the selected router CPU Utilization chart see CPU Utilization Drill Down on page 3 30 Memory Utilization chart see Memory Utilization Drill Down on page 3
159. erface for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 QuickView The QuickView report comprises the following components The Daily Grade of Service stacked bar chart shows the GOS scores for the selected interfaces The chart shows weighted values for the metrics for the report day For LAN interfaces the metrics are as follows Utilization Errors and Discards For WAN interfaces the metrics are as follows Input Utiliza tion Output Utilization Total Percent Errors and Total Percent Discards For a description of how GOS is calculated for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively Daily Grade of Service LAN Interfaces With Highest Grade of Service GOS Scores Mon Dec 20 1999 Critical Warning Excellent B Err M Discards M thilization BPI asofues wWlESy asofues SPIES EE3YPONn PIES HEPOD ney nBunUnosoy uopsnoy nooy OFS uognoy S29INO B EI0dI09 sinjues esin aay yueasay ESIN La Report Descriptions ReportPack Guide Double click on an interface in the chart update the following components with performance data for the selected interface LAN only The Hourly Volume bar chart shows the volume of data flow for each sample period of the report day for the selected LAN interface An overlay line graph shows the volume computed for each hour of the over the baseline period so you can compare the vo
160. erties as a tab delimited file Importing Property Data Property import is handled automatically by the system if the files are present in the designated SourceDirectory To import property data before the scheduled nightly process 1 From a command prompt change directory to the DPIPE_HOME scripts directory 2 Locate the TREND procedure file and run the procedure file using the following command trend_proc f lt ReportPackName gt _importdata pro This will truncate the existing property data table and import the new property data into the property data table using ee_collect The update script will update the property columns in the key table only if an entry exists in the property data table that has the same keyid value in both data table and key table 1 24 Using the Property Import Feature ReportPack Guide Removing the ReportPack D fe U D 2 gt 2 3 2 D 1 Note You cannot remove a ReportPack if other ReportPacks have a depen dency on that ReportPack To remove the ReportPack 1 Click the Start button and then point to Programs Point to the folder that contains TREND and then click TREND The TREND window is displayed Removing the ReportPack 1 25 1 ReportPack Administration 2 3 Click Autopilot The Autopilot dialog box is displayed Cisco_Router_Demo lt ersion 3 5 2 gt FrameRelay Switch Demo lt Version 1 0 gt System lt Ve
161. es from four tables in the Empire MIB and one MIBlet from the MIB II MIB See the TREND User s Guide for more information about MIBlets _h N ie lt 77 g Po G xo fe Data Source 12 33 y O x 12 System ReportPack Table 12 1 lists the variables from the kernelperf table in the empire mib file Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String cpul Min 1 3 6 1 4 1 546 1 1 7 8 1 The percentage of time over the last minute the CPU was not idle The agent continuously samples the CPU and updates this variable On multi processor machines this object represents the value for the first CPU only The Host Resources MIB hrProcessorTable breaks out the CPU load on a per processor basis This value is meaningless until at least two samples have been taken On NT systems this value is the average over all CPUs cpuSMin 1 3 6 1 4 1 546 1 1 7 8 2 The percentage of time over the last 5 minutes the CPU was not idle The agent continuously samples the CPU and updates this variable On multi processor machines this object represents the value for the first CPU only The Host Resources MIB hrProcessorTable breaks out the CPU load on a per processor basis This value is meaningless until at least two samples have been taken On NT systems this value is the average over all CPUs cpulSMin 1 3 6 1 4 1 546 1 1 7 8 3 The percentage of time over the last
162. eskTalk Systems Inc bufferMdSize 1 3 6 1 4 1 9 2 1 22 Contains the size of medium buffers bufferMdTotal 1 3 6 1 4 1 9 2 1 23 Contains the total number of medium buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferMdFree 1 3 6 1 4 1 9 2 1 24 Contains the number of free medium buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferMdHit 1 3 6 1 4 1 9 2 1 26 Contains the number of medium buffer hits Changed from INTEGER to Counter by DeskTalk Systems Inc 2 of 5 Data Source 8 31 D o S e U o i x J21NOY ODSID g 8 Cisco Router ReportPack 8 32 Table 8 1 Cisco Router Variables Variable OID String buffermdMiss 1 3 6 1 4 1 9 2 1 27 Contains the number of medium buffer misses Changed from INTEGER to Counter by DeskTalk Systems Inc bufferB gSize 1 3 6 1 4 1 9 2 1 30 Contains the size of big buffers bufferB gTotal 1 3 6 1 4 1 9 2 1 31 Contains the total number of big buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferB gFree 1 3 6 1 4 1 9 2 1 32 Contains the number of free big buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferB gHit 1 3 6 1 4 1 9 2 1 34 Contains the number of big buffer hits Changed from INTEGER to Counter by DeskTalk Systems Inc bufferB
163. esponse Time wasp 1 Thu Oct 21 12 00 AM 100 00 75 00 50 00 milliseconds 25 00 o 00 00 02 00 04 00 06 00 08 00 10 00 12 00 Figure 3 43 Service Level Hourly Response Time Drill Down System swioday wo gt Ss lt S gt Ko J m z oO Service Level Management Report 3 47 3 Analyzing TREND Reports TREND 3 48 TREND 4 Frame Relay ReportPack The Frame Relay ReportPack provides comprehensive data collection and reporting to support the management of Frame Relay networks It includes compound reports for capacity planning service level reporting forecasting troubleshooting and also provides an executive summary of the performance of all Permanent Virtual Circuits PVCs and reports the detail performance metrics for PVC networks The Frame Relay ReportPack answers questions such as Are the Frame Relay networks providing an acceptable level of service to the customers end users or both What PVCs are overloaded and are there underutilized PVCs that can help balance the load Are performance problems due to high levels of congestion errors excessive utilization or other causes What are the top ten PVCs contributing to the poor health of the network Which PVCs are likely to degrade service in the near future if preventive action is not taken The Frame Relay ReportPack collects data from the RFC 1315 MIB and some pro prietary MIBs which are supported
164. et is added to the divisor to prevent a divide by zero condition Percent Input Discards ifindiscards Gfinucastpkts ifinnucastpkts ifindiscards ifinerrors inunknownprotos 1 100 0 The ratio of the number of inbound packets that were discarded for reasons other than containing errors to all inbound packets regardless of type plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Percent Output Discards ifoutdiscards foutnucastpkts ifoutucastpkts 1 100 0 The ratio of the number of outbound packets that were discarded for reasons other than containing errors to the number of packets that higher level protocols requested be transmitted to a unicast or non unicast address plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Volume inoctets outoctets The inoctets and outoctets bytes for the selected WAN interfaces are totaled for the specified aggregation period Total Errors ifouterrors ifinerrors inunknownprotos Basic Metric Calculations ReportPack Guide Table 5 3 WAN Metrics syoequoday Metric Computed As o m gt z D a gt zZ The sum of outbound packets that contain errors inbound packets that contain errors and inbound packets from unknown protocols Total Discards ifindiscards ifoutdiscards The sum of i
165. eviously negative condition Report Descriptions ReportPack Guide Routers With Greatest Change In Discard Rate Thu Oct 21 1999 Device Pct Pkts Previous Pct Change Rank Rank Prior Rank Pkts sec Bytes sec Discarded Pkts Discarded cotati 2 52 2 63 1 6 6 4015947 52 174640564 83 lincoln 3 73 3 74 2 1 1 33926123 71 443368212 95 boston 1 86 1 85 3 8 8 3774174727 1213756965 47 sanluis 3 48 3 48 4 4 4 27366425 47 853449719 02 miami 3 73 3 73 5 2 2 27480781 14 1182570806 32 sanjose 3 73 3 73 6 3 3 27538522 93 1182195575 79 houston 2 58 2 58 7 5 5 11812231 66 972845001 10 tulsa 2 48 2 48 8 7 7 36379888 04 947004189 12 Service Level Management Report o gt D fe c Ema D D Ss e U o P x The Service Level Management report let CIOs CFOs network managers users and customers know if their routers are performing against contracted service levels It reports the ten routers with lowest availability and the ten routers with the highest network response times It also provides drill down reports for daily availability and daily and hourly response times for the routers with lowest availability or highest network response time The Service Level Management Report comprises the following components Report Descriptions 6 17 6 Router ReportPack The Routers with Lowest Availability table lists the routers with the lowest availability for the report day Availability Routers With Lowest Availability T
166. ew and Snapshot Reports ReportPack Guide Memory Utilization Drill Down Figure 3 28 shows the Memory Utilization line graph which displays memory utilization for each sample taken during the report period for the selected router Memory Utilization sanluis 1 Mon Dec 20 12 00 AM Mon Dec 20 10 45 PM 100 00 80 00 60 00 rercem 40 00 20 00 60 00 06 00 12 00 18 00 Figure 3 28 Near Real Time Memory Utilization Drill Down Bay Router swioday wo gt Ss lt S gt Ko J m z s Near Real Time QuickView and Snapshot Reports 3 31 3 Analyzing TREND Reports Buffer Utilization Drill Down Figure 3 29 shows the Buffer Utilization line graph which displays the buffer utilization for each sample taken during the report period for the selected router Buffer Utilization sanluis 1 Mon Dec 20 12 00 AM Mon Dec 20 10 45 PM 100 00 80 00 60 00 Percent 40 00 20 00 0 00 00 06 00 12 00 18 00 Figure 3 29 Near Real Time Buffer Utilization Drill Down Bay Router Hot Spots Report The Hot Spots report identifies areas where threshold values have been exceeded for example excessive utilization or error counts Intelligent exception thresholds are predefined in TREND but you can customize them to suit your enterprise For example the following components comprise the Hot Spots report for the Cisco Router ReportPack Problem Summary for the Day table see Problem Summar
167. faces or all Routers The report is structured for the corporate executive who can review the daily volume summary and if desired look into more detailed information such as Grade of Service GOS hourly volume compared to baseline and total exception count The intent of the report is to provide executives with insight into the usage and return on investment of their network resources For example the following components comprise the LAN Connectivity Executive Summary Daily Volume chart see Daily Volume Chart on page 3 3 Hourly Volume chart see Hourly Volume and Hourly Baseline Charts on page 3 4 Hourly Baseline chart see Hourly Volume and Hourly Baseline Charts on page 3 4 Hourly Grade of Service chart see Hourly Grade of Service Chart on page 3 5 Executive Summary Report ReportPack Guide Hourly Bandwidth Utilization chart see Hourly Bandwidth Utilization Graph on page 3 6 The following sections describe these charts in more detail For complete information about the LAN Connectivity ReportPack see LAN and WAN ReportPacks on page 5 1 Daily Volume Chart Figure 3 1 shows the Daily Volume bar chart which compares yesterday s total volume of traffic to the daily volumes observed over the baseline period which is the 42 days from November 9 through December 19 in this example In this example yesterday is December 19 a8 This report provides CFOs CIOs and
168. faces With Highest Exception Counts hion Dec 20 1999 Device Description Utilization Error Discard Total Exceptions Exceptions Exceptions Exceptions sanjose hiktgs 0 00 23 00 0 00 23 00 sanluis Corporate Offices 0 00 21 00 0 00 21 00 miami Sales 0 00 19 00 0 00 19 00 miami hiktgs 0 00 19 00 0 00 19 00 boston Northeast Sales 1 00 3 00 1 00 5 00 houston Stock Room 3 00 0 00 0 00 3 00 Following are the exceptions and thresholds for the LAN and WAN interfaces e LAN Utilization 40 Errors 5 and Discards 5 WAN Input Utilization 80 Output Utilization 80 In Errors 5 Out Errors 5 In Discards 5 and Out Discards 5 Double click on any interface listed in the table to update the following charts and graphs with performance data for the selected interface The Grade of Service stacked bar graph shows GOS scores for the LAN metrics Utilization Errors and Discards and WAN metrics Input Utilization Output Utilization Total Percent Errors and Total Percent Discards over the baseline period for the selected interface For more information about the GOS weighting calculations for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively Report Descriptions ReportPack Guide a a gt 32 2 553 Grade of Service 2 2 sanjose Sales Mon Dec 20 12 00 Aut 7 gt Critical Warning Good Excellent Mon 00 00 Mon03 00 Mon06 00 Mon09 00 Mon 12 00 Mon 15 0
169. ffer utilization for the router reaches the defined thresholds and the projected 95th percentile CPU memory and buffer utilization for the router in 90 days Estimated Days To Threshold Bay Routers Within 90 Days of Threshold Thresholds CPU Utilization 80 Memory Util 75 Buffer Util 75 Router Slot DTT CPU Util DTT Memory OTT Buffer 90 Day CPU 90 Day Memory 90 Day Buffer Forecast Util Forecast Util Forecast Util Forecast Util Forecast Util Forecast sanluis 1 N A 1000 00 NIA 91 96 100 00 98 75 sanjose 3 N A 1000 00 N A 91 39 100 00 98 75 sanjose 1 1000 00 541 00 1000 00 94 27 98 45 95 87 sanluis 3 1000 00 1000 00 1000 00 85 98 91 90 85 25 Double click on a Bay Router in the summary table to update the content of following components with performance information for the selected router The Bay Router Daily GOS stacked bar chart shows weighted values for CPU memory and buffer utilization for each day of the baseline period 7 6 Report Descriptions ReportPack Guide for the selected router See Table 7 3 for a description of the GOS weighting factors Bay Router Daily GOS sanjose 5 Mon Now 08 1999 Sun Dec 19 1999 Critical Warning Good Excellent Nov 08 Nov 22 Dec 06 E CPU Utilization E Buffer utilization Bitemory Utilization The Bay Router Daily CPU Utilization line graph shows the average and 95th percentile CPU utilization for the selected router for each day of the baseline per
170. fied in octets For interfaces that are used for transmitting network datagrams this is the size of the largest network datagram that can be sent on the interface ifSpeed 1 3 6 1 2 1 2 2 1 5 An estimate of the interface s current bandwidth in bits per second For interfaces which do not vary in bandwidth or for those where no accurate estimation can be made this object should contain the nominal bandwidth ifPhysAddress 1 3 6 1 2 1 2 2 1 6 The interface s address at the protocol layer immediately below the network layer in the protocol stack For interfaces which do not have such an address e g a serial line this object should contain an octet string of zero length ifAdminStatus 1 3 6 1 2 1 2 2 1 7 The desired state of the interface The testing 3 state indicates that no operational packets can be passed 2 of 5 Data Source ReportPack Guide a Dr 2 p gz Table 5 1 LAN and WAN Connectivity Variables 3 2 Variable OID String z Z ifOperStatus 1 3 6 1 2 1 2 2 1 8 The current operational state of the interface The testing 3 state indicates that no operational packets can be passed ifLastChange 1 3 6 1 2 1 2 2 1 9 The value of sysUpTime at the time the interface entered its current operational state If the current state was entered prior to the last re initialization of the local network management subsystem then this obj
171. fined for each ReportPack For more information about the property file formats for the bundled ReportPack see Property File Formats on page A 1 You must place the property file into the designated SourceDirectory the Re portPack will process the file from this directory TREND uses a configuration file called a TEEL file to determine how to find property import files The ReportPack uses the TEEL file to locate a new property file or files nightly and imports the con tents of those found It overwrites existing property assignments and insert new property assignments The TEEL file for each ReportPack is installed into the DP IPE_HOME 1ib folder by Autopilot during the ReportPack s installation The TEEL file has the following file name format Using the Property Import Feature 1 17 1 ReportPack Administration lt ReportPack gt _Property teel For example the TEEL file for the Frame Relay ReportPack would be Frame_Relay_Property teel Note Although the examples in the TEEL file show that the SourceDirectory resides in the DPIPE_HOME packages directory DeskTalk recom mends that you create and use a new directory instead This will avoid any possible problems when you install future ReportPack upgrades in the DPIPE_HOME packages directory If you change the SourceDirectory location as DeskTalk recommends you have to edit the ReportPack s TEEL file to define the SourceDirectory and SourceDisp
172. fixed maximum number of files that can be written to the device The value of this object indicates the maximum number of files that can be written to this device devFfiles 1 3 6 1 4 1 546 1 1 1 7 1 8 The number of free or unallocated files that can be created on this device devMaxNameLen 1 3 6 1 4 1 546 1 1 1 7 1 9 2 of 4 12 48 Data Source ReportPack Guide Table 12 5 System Variables Polled from Empire devTableEntry Table Variable OID String The maximum file name length that this mounted device supports If no fixed limit exists 0 is returned devType 1 3 6 1 4 1 546 1 1 1 7 1 10 The type of the mounted device If this underlying device does not support feature 0 is returned devFsid 1 3 6 1 4 1 546 1 1 1 7 1 11 The possibly unique file system ID of the mounted device V 4 systems support its uniqueness while SunOS does not For devices or operating systems not supporting this object the object instance identifier is returned instead devFstr 1 3 6 1 4 1 546 1 1 1 7 1 12 Some devices support a description string If supported by the underlying device and operating system that device specific description string is returned If not supported the string Not supported is returned An example description string may be User home directories devUnmount 1 3 6 1 4 1 546 1 1 1 7 1 13 Writing a delete 1 to this object causes the agent
173. formation you need for decision and action Note that not all reports have drill down analysis If you view the report with the TRENDview PRO Client you can select a row in the summary chart by drilling down double clicking on it which high lights the selected row in yellow The content of the drill down components changes to reflect the metrics for the selected summary row in this case the Houston router For instructions on using the TRENDview PRO Client see the Guide To Viewing TREND Reports Report Layout 2 3 2 Overview Standard Reports A report consists of a summary chart and one or more detailed drill down charts graphs or tables that support the summary data Table 2 1 lists the eight standard reports each TREND ReportPack provides Table 2 1 Standard Reports Report Description Executive Summary Provides high level graphs that aggregate key metrics for all managed elements within a technology for example all frame relay PVCs or all routers Executives can review the daily volume summary and if desired look into more detailed information such as grade of service hourly volume compared to baseline and total exception count Hot spots Identifies specific areas of concern where threshold values have been exceeded for example excessive utilization or error counts Managed elements in this report are ranked by total number of exceptions Each exception is linked to multiple drill down c
174. g usage from dsi_dbstats which is Utilization equivalent to the ratio of total bytes used to the size of the transaction log multiplied by 100 Database kbused The total number of bytes used by Usage the database a o Z Database dbsizekb The total number of bytes allocated g g Size for the database w 5 OD Transaction kblogused The total number of bytes used by a S Log Usage the transaction log 9 mo Transaction logsizekb The total number of bytes allocated P Log Size for the transaction log Basic Metric Calculations 11 11 11 TREND Database ReportPack TREND 11 12 TREND 12 System ReportPack The System ReportPack reports statistics about CPU utilization paging and swapping file system usage virtual and physical memory cache hits and misses and grade of service specifically for the Empire SystemEDGE agent The System ReportPack answers questions including Do I need to add more memory to my system Am Trunning out of filesystem space Is my system running low on resources Report Descriptions The following sections describe each report in the System ReportPack in detail 12 1 _ N w lt P 3 D O U o Q 12 System ReportPack Executive Summary Report The Executive Summary report provides high level graphs aggregating key metrics for all systems It allows executives to review the daily volume summary and if
175. gement Executive Summary Daily Volume for P Cs Thu Now 11 1999 Wed Dec 22 1999 97 00 77 60 58 20 Gigabytes 38 80 19 40 0 Nov 11 Nov 15 Nov19 Nov 23 Nov 27 DecO1 Dec 05 DecO9 Dec 13 Dec1 Dec21 The Hourly Volume bar chart shows the total volume for all PVCs for each hour of the report day Volume Baseine Hourly Volume All P Cs Wed Dec 22 1999 4 836 3 626 3 2 2 426 2 D 1 216 o 00 00 06 00 42 00 18 00 A 7 o 3 o v E 2o lt D Report Descriptions 4 3 ge 2 4 Frame Relay ReportPack The Hour of Day Baseline bar chart shows the average volume for each hour of the day over the baseline period You can compare total volume for each hour of the report day with the baseline for that hour for the report period Volume Baseline Hour of Day Baseline All PVCs Last 42 Days 4 186 3 136 a m 2 096 _ 1 046 0 00 00 06 00 12 00 18 00 The Hourly Grade of Service stacked bar chart provides an overview of PVC performance providing weighted values for utilization and congestion Table 4 3 describes how Grade of Service is calculated Hourly Grade of Service All PVCs Wed Dec 22 1999 Critical Warning Good Excellent Wed 00 00 Wed 09 00 Wed 18 00 E Congestion E out utilization W in Utilization 4 4 Report Descriptions ReportPack Guide The Hourly Utilization line graph calculates the average In Utilization and Ou
176. h sample taken on the report day Percent Errored Packets miami Acme s Main Office Thu Oct 21 1999 0 00 00 06 00 12 00 18 00 00 00 Report Descriptions 6 13 6 Router ReportPack The Hourly Volume bar chart shows the volume of data flow for each hour of the report day for the selected router Hourly Volume miami Acme s Main Office Thu Oct 21 1999 4 826 00 3 860 80 2 895 60 1 930 40 w a T a m gt a 965 20 0 Thu 00 00 Thu 09 00 Thu 18 00 The Packets Per Second line graph shows Throughput packet second for the selected router for each sample period on the report day Packets Per Second miami Acme s Main Office Thu Oct 21 1999 4 664 544 50 3 731 635 60 2 798 726 70 1 865 817 80 Packets Second 932 903 90 0 00 00 06 00 12 00 18 00 00 00 6 14 Report Descriptions ReportPack Guide Top Ten Report The Top Ten report lets the network manager know at a glance which routers have the highest throughput highest change in throughput highest discard rate and highest change in discard rate This report shows the routers with the greatest throughput that could present potential problems in the future It ranks the top ten routers by total throughput and discard rate as well as by highest change of rank Volume and health change charts alert users to the routers greatest changes in volume suggesting potential capacity overloading if increasing trends continue The volume
177. he device The Daily Response Time line graph shows the average response time for the selected PVC for each day of the baseline period For more information about how daily response time is calculated see Daily Response Time Drill Down on page 3 46 Report Descriptions 4 23 A TI o 3 D D o o lt D me fe 1 4 Frame Relay ReportPack Daily Howy Daily Response Time COPBKR_BCI01 Sales Thu Now 11 1999 Wed Dec 22 1999 251 20 188 40 125 60 milliseconds 62 30 0 Nov 11 Nov 18 Now25 DecO2 DecO9 Dec 16 The Hourly Response Time line graph shows the response time for each sample for the report day for the selected PVC For more information about hourly network response time see Hourly Response Time Drill Down on page 3 47 Daily Hourly Hourly Response Time COPBKR_BCI01 Sales Wed Dec 22 12 00 AM 460 00 337 50 225 00 milliseconds 112 50 0 00 00 02 00 04 00 06 00 08 00 4 24 Report Descriptions ReportPack Guide Data Source Source data for the Frame Relay ReportPack is collected from the Frame Relay MIB RFC 1315 which is supported by most routers CSU DSUs and other network devices Table 4 1 lists the variables that are polled by SNMP from the RFC 1315 MIB Table 4 1 Frame Relay Variables RFC 1315 Variable OID String frCircuitState 1 3 6 1 2 1 10 32 2 1 3 Indicates whether the particular virtual circuit is
178. he percentage of the daily average and maximum database utilization for the last 90 days Database Utilization Average and Maximum Last 90 Days Wed Apr 28 1999 Mon Jul 26 1999 85 80 Percent Utilized 0 Apr 28 May12 May26 Jun09S Jun23 JulO 7 Jul 21 Daily Average Utilization ULTRA_SY BASE 0 Daily Maximum Utilization ULTRA_SY BASE 0 Report Descriptions ReportPack Guide The Database Average Utilization bar chart shows the average percentage of the database used for the baseline period and the projected percent usage of the database for the 30 60 and 90 day forecasts Database Average Utilization 30 60 and 90 Day Forecast Baseline Tue Jun 15 1999 Mon Jul 26 1999 Percent Utilized 00 00 00 Wi baseline ULTRA_SYBASE 0 30 Day ULTRA SYBASE 0 60 Day ULTRA SYBASE 0 90 Day ULTRA SYBASE 0 The Transaction Log Utilization line graph shows the percentage of the daily average and maximum transaction log utilization for the last 90 days Transaction Log Utilization Average and Maximum Last 90 Days Wed Apr 28 1999 Mon Jul 26 1999 a 36 85 80 2o U Q H D 5 57 20 xo zr a o o p amp 28 60 5 W 0 Apr 28 May12 May26 Jun09S Jun23 Jul OF Jul 21 Daily Average Utilization ULTRA_SY BASE 0 Daily Maximum Utilization ULTRA_SY BASE 0 Report Descriptions 11 3 11 TREND Database ReportPack The Transaction Log Max Utilization bar chart shows the average per
179. hich shows the percentage of samples taken during the day in which the In Utilization swioday wo gt Ss lt S gt Ko J m z oO QuickView and Snapshot Reports 3 25 3 Analyzing TREND Reports metric calculated for the samples fall in a defined range For more information about frequency distribution see Frequency Distribution on page 2 14 Distribution of Bandwidth Utilization Input Sanjose Support Wed Dec 22 1999 o 20 i 21 40 41 60 f meio f 081 100 P EMES Figure 3 24 Quick View Distribution of Bandwidth Input Distribution of Bandwidth Utilization Output Figure 3 24 show Distribution of Bandwidth Utilization Output pie chart which displays the percentage of samples taken during the day in which the Out Utilization 3 26 QuickView and Snapshot Reports ReportPack Guide metric calculated for the samples fall in a defined range For more information about frequency distribution see Frequency Distribution on page 2 14 Distribution of Bandwidth Utilization Output Sanjose Support Wed Dec 22 1999 mo 20 L 21 40 41 60 f Wict s0 f Os1 100 miy Figure 3 25 Quick View Distribution of Bandwidth Output Near Real Time QuickView and Snapshot Reports suod y The Near Real Time QuickView and the Near Real Time Snapshot reports provide device performance statistics up to the last SNMP poll The reports does not rely on nightly summaries th
180. his file manually you will need to provide them LAN WAN Connectivity ReportPack Table A 1 lists the property file formats for the LAN WAN Connectivity ReportPack Table A 1 LAN WAN Connectivity ReportPack Property File Formats customer Column Name Function Value Type dsi_target_name Device name whichis the name varchar 64 you use to poll the device dsi_table_key The device index which is the varchar 128 ifindex value for the interface IfSpeed The interface speed in bits sec float ond The ReportPack requires this value cust_id Unique integer value for each integer A 2 LAN WAN Connectivity ReportPack ReportPack Guide gt nz Y Table A 1 LAN WAN Connectivity ReportPack Property File Formats 3 0 D g lt Column Name Function Value Type at o customer_name Customer name associated with varchar 128 cust_id region_id Unique integer value for each integer region region_name Region name associated with varchar 128 region_id value dsi_descr A description of the interface varchar 128 Frame Relay ReportPack Table A 2 lists the property file formats for the Frame Relay ReportPack Table A 2 Frame Relay ReportPack Property File Formats Column Name Function Value Type dsi_target_name Device name which is the name varchar 64 you use to poll the device dsi_table_key The device index which is the varchar 128
181. hold value Forecasting ReportPack Guide A default threshold value is supplied with each TREND ReportPack which you can change to suit your installation TRENDsum forecasting calculations are based on a least squares linear regression which means that the calculations attempt to draw a straight line through a set of data points collected over time as shown in Figure 2 2 N ie lt Time p Figure 2 2 Forecasting Least Squares Linear Regression Each data point represents a daily 95th percentile value for the element property being forecast The 95th percentile is the value below which 95 of the samples taken during the day fall You want to look at peak utilization for planning purposes so a daily 95th percentile value which is a near maximum value is used rather than a simple average The 95th percentile better characterizes the shape of the data by taking into account the measurement variations spikes that occur during the day averaging on the other hand tends to smooth out the effect of measurement variation thereby giving a false impression of capacity For a DTT calculation you have the following cases Forecasting 2 9 2 Overview You will hit the threshold in the calculated number of days You are already beyond the threshold a negative value You are never going to hit the threshold the line slopes downward away from the threshold value You are a long distance
182. hour for the interface selected from the Segment Utilization table The Frame Size Distribution stacked bar chart shows the total number of frames and the distribution of frame size by hour for the interface selected from the Segment Utilization table D Ed e Z N D fe V o Q Snapshot Reports The RMON2 ReportPack provides the following Snapshot reports Daily Snapshot Hourly Snapshot They provide the same information as their corresponding Quick View reports however you can select the specific interfaces that you want the report to include For more information about selecting an interface for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 Top Ten Report The Top Ten reports list the top ten conversation pairs talkers and listeners for the preceding day and preceding week Specifically it lets you determine if your interfaces are not optimally deployed have a hardware problem or overloaded by requests Report Descriptions 10 9 10 RMON2 ReportPack Daily Top Ten The Daily Top Ten report lists the top ten conversation pairs talkers and listeners for the preceding day The Daily Top Ten report comprises the following components The Top Conversation Pairs table lists the source and destination IP address pairs that sent the most data on the preceding day along with the total megabytes sent Top Converstion Pai
183. hu Oct 21 1999 Device Availability sanluis 100 00 miami 100 00 tulsa 100 00 cotati 100 00 lincoln 100 00 boston 100 00 houston 100 00 Double click a router in the table to update the following graphs with performance data for the selected router The Daily Availability line graph compares the availability of the selected router with the Service Level Agreement over the baseline period By default the service level is set at 99 5 Daily Hourly Daily Availability sanjose Fri Sep 10 1999 Thu Oct 21 1999 100 00 80 00 z 60 00 3 F a 40 00 20 00 0 Sep 10 Sep17 Sep24 OctO1 OctO8 Oct 16 Availability Service Level 6 18 Report Descriptions ReportPack Guide The Hourly Availability line graph compares the availability of the selected router with the Service Level Agreement over the baseline period for the report day By default the service level is set at 99 5 Daily Hourly Daily Availability sanjose Thu Oct 21 12 00 AM Fri Oct 22 12 00 AM 100 00 80 00 60 00 Percent 0 4 00 00 03 00 06 00 09 00 12 00 16 00 18 00 21 00 00 00 Availability Service Level The Routers with Highest Response Time table lists the routers with the highest average response times for the day in descending order Response Time Routers With Highest Response Time Thu Oct 21 1999 Device Response Time miami 500 00 boston 479 52 sanluis 452 32 cot
184. ice GOS Sanjose Support Thu Nov 11 1999 Wed Dec 22 1999 Critical Warning E Congestion E Out utilization Win utilization Good Excellent Nov 11 Now15 Nov19 Nov23 Nov27 DecO1 DecO5 DecO9 Dec13 Dec1 Dec21 Figure 3 8 Forecast Grade of Service Drill Down Frame Relay Health scores for each metric are assigned and weighted according to their relative importance added together with other metric scores and a final grade is automati cally calculated If the default polling interval is used samples are taken every 15 minutes so a daily Grade of Service GOS grade represents the average of the weighted composite scores for 96 daily samples By transforming numerous service related statistics into an easily understood stacked bar graph and presenting it over the baseline period the GOS chart helps you recognize changes in Frame Relay PVC health and the root cause of those changes The Frame Relay GOS metrics are Input Utilization Output Utilization and Congestion Utilization is expressed as a percent of CIR Committed Information 3 10 Forecast Report ReportPack Guide Rate if that value is available otherwise it is expressed as a percent of ifSpeed interface speed A health score is assigned to each metric A value of 1 corresponds to an Excellent rating Since not all metrics may impact service quality equally each metric is multiplied by a weighting factor For Frame Relay Input and Output Utilization e
185. igned Region lincoln 1 79 2 Belle Express 2 Unassigned Region orlando 1 30 2 Belle Express 2 Unassigned Region orlando 1 50 2 Belle Express 2 Unassigned Region 4 Ensure that the User ID value you assign using the TRENDweb User Adminis tration application matches the cust_id value that you assigned in the prop erty datapipe for each customer For more information about how to create user accounts see the TRENDweb Installation and Administration Guide Note If the cust_id does not match the value of User ID then you may not see your reports 5 Import the property file using the instructions in Importing Property Data on page 1 24 Note If a new PropertyImport changes existing CustomerID assignments TREND will begin using the new assignments immediately for all subsequent aggregations However aggregations that were previously processed will not be reprocessed based on the new assignments Exporting Data To export data you must generate a property file containing the data you want the importing process to use by doing either of the following Using the Property Import Feature 1 21 1 ReportPack Administration 1 22 Use the property datapipe to generate the property file s Create your own property files Generating Property Files To generate a property file 1 From a command prompt change directory to the DPIPE_HOME scripts directory Locate the ReportPack s proc
186. ii Preface 2 ii bees seid ee eee idee i ee ieee de eee ieee teen xV Organization of This Guide 0 2 62 eee XV Related Documents 000 00 cee eee eens xvii Online User Manual Availability 0 0 0 0 eee ee xvii ReportPack Administration 00 cece eee eee eee eee 1 1 ReportPack DependencieS 0 0 cence eee eee 1 2 Installing ReportPacks 0 0 cece ee 1 2 Unzipping the ReportPack Sources 00 0c 1 3 ReportPack Directory Structure 0 20000 eee eaee 1 4 Types of ReportPack Installations 0 5 1 5 New Installation 0000 00 cece ee ees 1 5 Upgrade Installation 2 0 2 0 00 0 cee eee 1 5 Upgrading from a V3 4 x RMON or RMON2 ReportPack 1 6 Running the RMON2 ReportPack Upgrade Procedure 1 7 Troubleshooting the RMON2 Upgrade Procedure 1 9 Demo Installation 002000 00 cee ee 1 9 Contents Running Autopilot RER EESSI eee 1 10 Using the Property Import Feature 0 000 ee eee eee 1 14 How Property Import Works 0 0000s 1 17 Defining the SourceDirectory anaana eee eee 1 18 Defining the SourceDisposition 0200 eee 1 19 Sample Property File 0 00 e cece ee eee eee 1 19 Displaying Data for Specific Customer Devices in Reports 1 20 Exporting Data 0 redea ersa a ad a ee 1 21 Generating Property Files aaua 1
187. ing the day then the total is divided by the number of samples taken which results in a bucket total sample count ratio That ratio is then multiplied by 100 to give a percentage of samples for the day whose In Utilization value falls within that bucket Table 2 3 shows a sample calculation of frequency distribution for the In Utilization metric Table 2 3 Sample Calculation for Frequency Distribution Input Utilization Buckets Frequency Frequency Distribution 0 20 7 7 30 23 3 21 40 12 12 30 40 41 60 7 7 30 23 3 61 80 3 3 30 10 81 100 1 1 30 3 7 gt 100 0 0 30 0 Total 30 100 2 14 Frequency Distribution ReportPack Guide Figure 2 3 shows a sample of how TREND depicts frequency distribution Distribution of Bandwidth Utilization Input orlando 1 30 Sun Aug 30 1998 W0 20 orlando 1 30 21 40 orlando 1 30 41 60 orlando 1 30 61 80 orlando 1 30 81 10 orlando 1 30 E gt 100 orlando 1 30 N O lt Figure 2 3 Sample Chart for Frequency Distribution Frequency Distribution 2 15 TREND TREND Analyzing TREND Reports ReportPacks typically contain the following reports Executive Summary Hot Spots Forecast Capacity Planning Top Ten QuickView swioday Snapshot wo gt lt S gt Ko J m z oO Near Real Time gt gt Service Level M
188. ing to a specific customer This feature is only available for the following V4 0 Bundled ReportPacks LAN WAN Connectivity Router Cisco Router Frame Relay Bay Router This feature will be available for other selected ReportPacks at a later date Using the Property Import Feature ReportPack Guide Note You can use the property import feature after you install the Report Pack However use of the property import feature is not mandatory If you do not need to use it you can ignore the subsequent sections D D U D Q 2 gt 2 3 D 1 This feature is useful for service provider organizations who want to provide seg mented reporting to different users or customers For example suppose Thunderbolt Inc has offices in Boston Pittsburgh and in Los Angeles The TRENDweb Admin istrator can configure the key tables after the ReportPack installation so that users in the Boston Pittsburgh and Los Angeles offices will be treated as separate custom ers seeing only the devices belonging to their office Thus when a user displays an Executive Summary report the report only shows performance statistics for the de vices relevant to the user s office The user s office is displayed at the top of the Ex ecutive Summary report However when the TREND Administrator logged in as trendadm he would see all the offices shown under the Customer Name category and a special All categ
189. inunknownprotos ifoutnucastpkts ifoutucastpkts 1 100 0 This metric is essentially the ratio of the sum of inbound and outbound packets that are discarded for reasons other than containing errors to the total packet traffic inbound and outbound at the interface plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Volume inoctets outoctets The inoctets and outoctets bytes for the selected managed elements are totaled for the specified aggregation period 5 32 Basic Metric Calculations ReportPack Guide Table 5 3 lists the calculations for the performance metrics used in WAN reports Note that some charts modify these calculations syoequoday o m gt z D a gt zZ Table 5 3 WAN Metrics Metric Computed As Input Utilization ifinoctets 8 delta_time ifspeed 1 100 0 inoctets are multiplied by 8 to give the number of bits which is divided by delta_time to give bits per second Bits per second is divided by ifspeed which is collected as bits per second plus packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Output ifoutoctets 8 delta_time ifspeed 1 100 0 Utilization outoctets are multiplied by 8 to give the number of bits which is divided by delta_time to give bits per second Bits per second is divided by ifspeed which is collected as b
190. iod Bay Router Daily CPU Utilization N sanjose 5 ton Now 08 1999 Sun Dec 19 1999 KS lt 122 80 o uvo 92 10 1 aS o z cr a 61 40 JI a E 30 70 Q a Nov 22 Dec 06 Awg CPU Utilization P95 CPU Utilization Report Descriptions 7 7 7 Bay Router ReportPack The Bay Router Daily Buffer Utilization line graph shows the average and 95th percentile buffer utilization for the selected router for each day of the baseline period Bay Router Daily Buffer Utilization sanjose 5 tion Nov 08 1999 Sun Dec 19 1999 129 60 97 20 64 80 Percent 32 40 Nov 22 Dec 06 Awg Buffer Utilization P95 Buffer Utilization The Bay Router Daily Memory Utilization line graph shows the average and 95th percentile memory utilization for the selected router for each day of the baseline period Bay Router Daily Memory Utilization sanjose 5 tion Now 08 1999 Sun Dec 19 1999 130 80 or 98 10 65 40 Percent 32 70 Nov 22 Dec 06 Avg Memory Utilization P95 Memory Utilization 7 8 Report Descriptions ReportPack Guide QuickView and Snapshot Reports The QuickView report provides the network management staff with detailed information on key metrics for individual Bay routers with the highest GOS scores The Snapshot report provides the same information as the QuickView report for individual Bay routers that you select from a pick list for inclusion in the report For more
191. its per second plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Utilization ifoutoctets ifinoctets 8 delta_time ifspeed 1 100 0 combined Input and Output Utilization The sum of outoctets and inoctets is multiplied by 8 to give the number of bits which is divided by delta_time to give bits per second Bits per second is divided by ifspeed which is collected as bits per second plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Input Errors ifinerrors inunknownprotos ifinucastpkts ifinnucastpkts ifindiscards ifinerrors inunknownprotos 100 0 Basic Metric Calculations 5 33 5 LAN and WAN ReportPacks 5 34 Table 5 3 WAN Metrics Metric Computed As The ratio of the number of inbound packets containing errors ifinerrors plus the number of inbound packets from an unsupported protocol inunknownprotos to all inbound packets regardless of type is multiplied by 100 to yield a percentage Percent Output Errors ifouterrors ifoutucastpkts ifoutnucastpkts 1 100 0 The ratio of the number of outbound packets containing errors if outerrors to the number of packets that higher level protocols requested to be transmitted to a unicast or non unicast address is multiplied by 100 to yield a percentage In this computation a negligible amount that is 1 pack
192. ix weeks prior to and including the report date The Forecast report comprises the following components The Estimated Days to Threshold table ranks Cisco routers that are within 90 days of reaching a 95 CPU Utilization 95 Memory Utilization or 75 Buffer Utilization thresholds in ascending order For more information about the 95th percentile days to threshold DTT and 90 day forecast methods of calculation see the TRENDsum utility section in Appendix A of the TREND User s Guide Report Descriptions ReportPack Guide Estimated Days To Threshold DTT Cisco Routers Within 90 Days Of Threshold Chresholds CPU Utilization 95 Memory Util 75 Buffer Util 75 Router DTT CPU 90 Day CPU DTT Physical 90DayMem DTT Buffer Util 90 Day Buffer Utilization Util Forecast Memory Util Util Forecast Util Forecast boston 12 00 112 14 128 00 84 12 1000 00 44 37 lincoln 1000 00 48 20 545 00 90 41 NIA 59 02 houston 0 00 98 00 1000 00 98 36 1000 00 22 90 cotati N A 6 99 1000 00 98 27 1000 00 93 78 Double click a router in the Estimated Days to Threshold table to update the content of following components with performance data for the selected router The Daily Grade of Service stacked bar chart shows grade of service scores for CPU Utilization Memory Utilization Buffer Utilization and Buffer Faults for the selected Cisco router for each day of the baseline period Daily GOS tulsa Desktalk Support 4 Mon Nov 08 1999
193. ization are computed Overutilized P Cs Projected To Exceed 70 Input or Output Utilization Within 90 Days All Values Based On 95th Percentile Device Interface DLCI Description Projectedin Projected Out Projectedin Projected Out Projectec Util 30 Days Util 30 Days Util 60 Days Util 60 Days Util 90 Da portland 1 21 Accounting 186 61 186 30 199 93 199 16 213 26 torrance 1 89 Support 184 79 187 06 197 10 200 43 209 42 portland 1 30 Sales 185 73 186 24 198 38 199 10 211 02 torrance 1 90 Marketing 96 54 103 49 110 99 116 84 125 44 Sanjose 1 20 Support 95 33 104 05 109 08 117 71 122 82 torance 1 88 Sales 95 19 103 56 108 68 116 94 122 16 sanjose 1 40 Marketing 95 55 102 92 109 48 116 88 123 42 orlando 1 30 Accounting 65 36 64 72 79 63 78 75 93 89 LO 4 12 Report Descriptions ReportPack Guide The Underutilized PVCs table ranks the PVCs projected to be the least utilized in terms of In Utilization or Out Utilization within the next 90 days in ascending order Underutilized PVCs Projected To Be Least Utilized In 90 Days All Values Based On 95th Percentile P95 Device InterfaceDLCI Description CurrentP95In Projectedin Projectedin Projectedin Curr Utilization Util 30 Days Util 60 Days Util 90 Days U Sales COPBKR_BCIO1 202104 0 23 COPBKR_BCIO1 202104 0 24 Sales 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 27 Sales 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 29 Sales 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 22 Suppo
194. kTalk Systems Inc bufferHgFree 1 3 6 1 4 1 9 2 1 64 4 of 5 D o S e U o i x J21NOY ODSID 8 Data Source 8 33 8 Cisco Router ReportPack Table 8 1 Cisco Router Variables Variable OID String Contains the number of free huge buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferHgHit 1 3 6 1 4 1 9 2 1 66 Contains the number of huge buffer hits Changed from INTEGER to Counter by DeskTalk Systems Inc bufferHgMiss 1 3 6 1 4 1 9 2 1 67 Contains the number of huge buffer misses Changed from INTEGER to Counter by DeskTalk Systems Inc 5 of 5 Basic Metric Calculations 8 34 The performance metrics used in Cisco Router reports are based on the calculations listed in Table 8 2 However some charts modify these calculations so be sure to check the metrics given for each chart in each report in subsequent sections of this chapter Table 8 2 Cisco Router Metrics Metric Computed As CPU Utilization cpubusy5 The ratio of the amount of CPU time used collected as centiseconds and multiplied by 100 to yield seconds to the maximum amount of CPU time available collected in centiseconds and multiplied by 100 to yield seconds is multiplied by 100 to yield a percentage Basic Metric Calculations ReportPack Guide Table 8 2 Cisco Router Metrics Metric Computed As Memory Utiliza
195. l for the TREND database administrator that shows trends in the database use and also a means to predict when to increase the database capacity of the system The TREND Database ReportPack answers questions including Do TIneed to increase the size of the database for better utilization Do I have enough space to load new applications or ReportPacks The report consists of a combination of charts and graphs To produce the report the TREND Database ReportPack uses data from the TREND processes in their normal mode of operation There is no data collection for this report in the normal sense Report Descriptions The following section describes the report in the TREND Database ReportPack in detail E m oz T a p 2 O gQ zo ts o O e 11 TREND Database ReportPack Database Usage Summary Report The Database Usage Summary report enables the TREND administrator to view the historic utilization of the TREND database and transaction log as well as view forecasts for 30 60 and 90 days into the future The TREND administrator can use this information to better utilize disk space while TREND continues to collect data The forecasts use the data from the baseline period which by default is the previous 42 days or 6 weeks to predict the usage level for the next 30 60 or 90 days The Database Usage Summary report comprises the following components The Database Utilization line graph shows t
196. le 2 2 on page 2 6 For more information about Grade of Service as it applies to each type of report see Analyzing TREND Reports on page 3 1 Grade of Service Calculation 12 51 _ N w lt P 3 D O U Q 12 System ReportPack TREND 12 52 TREND Property File Formats This appendix describes the property file formats for the following ReportPacks LAN WAN Connectivity Frame Relay Router Cisco Router gt Bay Router Note The Cisco and Bay Router ReportPacks use the same property file formats as the Router ReportPack Ensure that you have the Router ReportPack installed before you use the Cisco or Bay Router ReportPacks You generate or create property files when you use the property export feature An example property file is provided for each ReportPack It is located in the following directory sewo gt Uv e me D gt lt at A Property File Formats DPIPE_HOME packages lt ReportPack gt lt ReportPack gt ap PropertyData Archive When you create your own property files ensure you adhere to the formats specified for each ReportPack For more information about the property import feature see Using the Property Import Feature on page 1 14 Note Do not change the dsi_target_name or dsi_table_key columns if you used the property export feature to create the data files If you are creating t
197. lity sanjose Mon Nov 08 1999 Sun Dec 19 1999 100 00 80 00 60 00 Percent Nov15 Now22 Nov29 DecO6B Dec13 Availability Service Level The Hourly Availability line graph shows the average hourly availability and service level shows for the selected Bay router for each sample for the preceding day Daily Hourly Hourly Availability sanjose Sun Dec 19 12 00 AM Mon Dec 20 12 00 AM 100 00 80 00 60 00 Percent 0 4 00 00 03 00 06 00 09 00 12 00 16 00 18 00 21 00 00 00 Availability Service Level 7 24 Report Descriptions ReportPack Guide The Response Time table lists the Bay routers with the highest average response times for the day Response Time Routers With Highest Response Time Sun Dee 19 1999 Device Response Time sanjose 386 141 Double click on a router in this chart to update the following components with performance data for the selected router The Daily Response Time line graph shows the daily response time for the selected Bay router for each sample over the report period 42 days Daily l Hourly l Daily Response Time sanluis Unassigned Mon Nov 08 1999 Sun Dec 19 1999 561 50 336 90 milliseconds 224 60 112 30 Nov 15 Nowv22 Nov29 DecOB Dec13 Report Descriptions 7 25 yoed N ius o lt D e 5 Po 1 xe e 1 7 Bay Router ReportPack The Hourly Resp
198. lity chart see Hourly Availability Drill Down on page 3 43 Response Time table see Response Time on page 3 45 and the following drill down components Daily Response Time chart see Daily Response Time Drill Down on page 3 46 swioday Hourly Response Time chart see Hourly Response Time Drill Down on page 3 47 i gt lt S Ss Ko J m z The following sections describe these charts in more detail For more information about the System ReportPack see System ReportPack on page 12 1 Service Level Management Report 3 41 3 Analyzing TREND Reports Availability Table Figure 3 38 shows the Availability table and its Daily and Hourly Availability drill down charts The Availability table lists the systems with the lowest availability TREND processes each system s availability results and highlights problem systems whose availability is below contracted Service Level Agreement SLA metrics TREND implements a management by exception policy by sifting through all managed systems and focusing your attention on only the problem units This reduces the amount of data you must review to stayed informed on system health Availability Systems With Lowest Availability Thu Oct 21 1999 Device Availability leyte 110 94 hornet 110 94 laffey 110 94 nimitz 110 94 yorktown 110 94 eisenhower 110 94 enterprise 110 94 Figure 3 38 Service Level Availability T
199. lume shown for the report day with the baseline Hourly Volume boston Northeast Sales Mon Dec 20 12 00 AM 487 50 Megabytes 0 00 00 06 00 12 00 18 00 LAN only The Distribution of Bandwidth Utilization pie chart shows the percentage utilization values that fall within the defined utilization buckets for the selected LAN interface for the report day Report Descriptions 5 7 syoequoday o m gt z D a gt z 5 LAN and WAN ReportPacks Distribution of Bandwidth Utilization boston Northeast Sales Mon Dee 20 1999 f 4 100 WAN only The Distribution of Input Bandwidth Utilization pie chart shows show the percentage input utilization values that fall within the defined utilization buckets for the selected WAN interface for the report day Input output Distribution of Input Bandwidth Utilization boston WAN Link Boston San Diego Mon Dec 20 1999 H0 35 In Util W 36 40 In Util 041 45 In Util E 46 100 In Util 5 8 Report Descriptions ReportPack Guide WAN only The Distribution of Output Bandwidth Utilization pie chart shows the percentage output utilization values respectively that fall within the defined utilization buckets for the selected WAN interface for the report day syoequoday o m gt z D a gt z Input Output Distribution of Output Bandwidth Utilizatio boston WAN Link Boston San Diego Mon Dec 20 1999 Wo
200. lyzing TREND Reports on page 3 1 8 36 Grade of Service Calculation o D Oo Z D UV D O TREND 9 RMON ReportPack The RMON ReportPack helps TREND users manage the various segments of their Ethernet networks Token Ring networks or both It helps identify possible trouble spots and pinpoint what may have caused the problem or point out a situation that may become a problem Remote Monitoring RMON is an IEFT standard SNMP Management Information Base designed to provide network performance information remotely Typically it is implemented by hardware or software that promiscuously reads packets off of the network scans their header information and collects a large number of statistics about traffic crossing the network These statistics include the number and size of packets traversing the network by protocol source and destination The RMON agent provides this information using SNMP and in some cases proprietary protocols The ReportPack provides four Quick View reports The reports help you identify net works that are overutilized have a high number of collisions or both These factors can indicate that a segment is overutilized A possible response to an overutilized segment is to create another segment A low utilization level combined with a high percentage of collisions could indicate a bad Ethernet card on the segment which would require replacing the card 9 1 9 RMON ReportPack U
201. managers users and customers know how their Frame Relay service is performing against contracted service levels It reports the ten PVCs with lowest availability and the ten PVCs with the highest network response times It also provides charts for daily availability and daily and hourly response times for the elements with lowest availability or highest network response time The Service Level Management Report comprises the following Report Descriptions ReportPack Guide The PVCs with Lowest Availability for the Day bar chart shows PVCs with the lowest availability for the previous day For more information about how availability is calculated see Availability Table on page 3 42 Availability PVCs With Lowest Availability Wed Dec 22 1999 Device Interface DCLI Availability portland 1 30 100 00 torrance 1 88 100 00 atlanta 1 24 100 00 atlanta 1 42 100 00 sanjose 1 40 100 00 torance 1 89 100 00 lincoln 1 19 100 00 orlando 1 30 100 00 torrance 1 90 100 00 portland 1 21 100 00 Double click on a device to update the content of the following drill down components with performance data for the device The Daily Availability line graph compares the availability of the selected PVC with the Service Level Agreement over the baseline period By default the service level is set at 99 5 For more information about how Report Descriptions 4 21 A TI o 3 D D o o lt D me fe 1 4 Frame Rel
202. mary report allows executives to review how Cisco Router networks as a whole are performing in the critical areas of volume CPU utilization memory utilization and buffer utilization over time Note The TRENDweb Administrator can configure the Cisco Router Report Pack installation so that the Executive Summary reports will only dis play data for devices belonging to a specific customer In Figure 3 1 the report displays information for only the devices belonging to the Acme company For more information about this feature see Displaying Data for Specific Customer Devices in Reports on page 1 20 The Executive Summary report comprises the following components 8 2 Report Descriptions ReportPack Guide The Cisco Router Inventory table shows all the Cisco routers used in the enterprise the number and types of interfaces LAN WAN or other and the number of routers managed Cisco Router Inventory Sun Dec 19 1999 Vendor WAH Interfaces LAN Interfaces Other Total Routers Managed Interfaces Interfaces Cisco 3 1 0 4 The Daily Volume bar chart shows the total volume for all Cisco routers for each day of the baseline period giving the executive an instant view of increases and decreases in Cisco router usage over time The default baseline period is the six weeks prior to and including the report date Hourly Daily Daily Volume All Cisco Routers Mon Nov 08 1999 Sun Dec 19 1999 40 80 30 60 20 40
203. memory utilization value All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Table 12 6 describes how memory utilization is computed Report Descriptions 12 23 _ N w lt P r 3 D O U Q 12 System ReportPack 12 24 Double click on a system in this table to update the following components with performance data for the selection The Grade of Service stacked bar chart shows GOS scores for CPU memory and swap utilization and queue length over the baseline period for the selected system See Table 12 7 for a description of the GOS weighting factors The Memory Utilization line graph shows the average memory utilization over each day of the baseline period for the selected system The CPU Utilization line graph shows the average CPU utilization over each day of the baseline period for the selected system The Filesystem Utilization line graph shows the average filesystem utilization over each day of the baseline period for the selected system The Swap Utilization line graph shows the average swap utilization over each day of the baseline period for the selected system The Page Faults bar chart shows the number of page faults for each day of the baseline period for the selected system The Underutilized System Memory summary table shows the systems that are projected to be the least utilized within the next 90 days ranked i
204. n ascending sequence by the 90 day projected memory utilization value All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Underutilized means that the system is furthest from reaching the utilization thresholds 30 defined Double click on a system in this table to update the following components with performance data for the selection The Grade of Service stacked bar chart shows GOS scores for queue length and CPU memory and swap utilization over the baseline period for the selected system See Table 12 7 for a description of GOS weighting factors The Memory Utilization line graph shows the average memory utilization over each day of the baseline period for the selected system The CPU Utilization line graph shows the average CPU utilization over each day of the baseline period for the selected system Report Descriptions ReportPack Guide The Filesystem Utilization line graph shows the average filesystem utilization over each day of the baseline period for the selected system The Swap Utilization line graph shows the average swap utilization over each day of the baseline period for the selected system The Page Faults bar chart shows the number of page faults for each day of the baseline period for the selected system Hot Spots Report The Hot Spots report identifies specific areas of concern where threshold values have been exceeded The top ten
205. n Dee 19 1999 GE cPu util 0 19 cpu util 20 39 CPU util 40 79 HE cP util 80 100 7 10 Report Descriptions ReportPack Guide The Memory Utilization Frequency Distribution pie chart shows the percentage of memory utilization values computed from each sample taken during the report day that fall within the defined utilization buckets for the selected Bay router Memory Utilization Frequency Distribution sanjose 1 Sun Dee 19 1999 Bntemory util 0 19 Dntemory Util 20 39 Otemory Util 40 79 Btemory util 80 100 The Buffer Utilization Frequency Distribution pie chart shows the percentage of buffer utilization values computed from each sample taken during the report day that fall within the defined utilization buckets for the selected Bay router Buffer Utilization Frequency Distribution sanjose 1 Sun Dee 19 1999 Bi buffer util 0 19 E Buffer util 20 39 E buffer Util 40 79 Wi buffer Util 80 100 Report Descriptions yoed N ius o lt D fo Po xe fo Di 7 Bay Router ReportPack The Percent Errors line graph shows the percent errors for the selected Bay router over the samples for each hour of the report day Table 7 2 explains how percent errors is computed Percent Errors sanjose Sun Dec 19 1999 6 30 5 10 Percent w 1 70 0 00 00 06 00 12 00 18 00 00 00 The Total Number of Discarded Packets line graph shows the total number of discar
206. n SunOS the kernel updates this parameter once every 5 seconds A large value may indicate the system does not have enough memory for the current load swapActive 1 3 6 1 4 1 546 1 1 7 8 7 The number of jobs that are active yet swapped out of memory sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds A high value for this object may indicate that system memory is not sufficient for the current load P sleepActive 1 3 6 1 4 1 546 1 1 7 8 8 Er oO 2 of 9 g J xo Q Data Source 12 35 S 12 System ReportPack Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The number of jobs that are active yet sleeping in core sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds A process is active yet sleeping in core when it is not running and not swapped out to secondary storage swap memInUse 1 3 6 1 4 1 546 1 1 7 8 9 The total real KBytes of memory in use sampled over a fixed period of time The sampling interval is system dependent In SunOS the kernel updates this parameter once every 5 seconds Comparing memInUse to memory in the Empire system group can help indicate if a system has sufficient memory for its current load activeMem 1 3 6
207. naged elements Highlights volatile elements as significant behavior changes are often indicators of service level problems in the making Lists current rank previous rank and change in rank which enables the network manager to assess the severity of network changes N fe lt p QuickView Displays the top ten problem elements ranked by GOS score and drill down charts appropriate to the technology for example utilization errors and volume compared to baseline Snapshot Includes the same information as the Quick View report but the element is selected from a pick list that is displayed when the report is invoked Enables the network manager to view the status of any element immediately not just the top ten problem elements Service Level Shows response time and availability plotted against Management contracted service levels Shows whether or not service levels have been achieved for the previous day month or other time period 2 of 2 Standard Reports 2 5 2 Overview For a selective description of the content of these reports for the various ReportPack technologies see Analyzing TREND Reports on page 3 1 Subsequent chapters provide detailed descriptions of the standard reports for each of the ReportPacks Metric Classes Table 2 2 lists report metrics One or more of the metrics can be included in each report Table 2 2 Metric Classes Metric Descri
208. nbound packets that are discarded and outbound packets that are discarded Percent Errors ifouterrors ifinerrors inunknownprotos ifoutucastpkts ifoutnucastpkts inunknownprotos ifinerrors ifinucastpkts ifinnucastpkts ifindiscards 1 100 0 The ratio of the sum of inbound and outbound packets containing errors and inbound packets from unknown protocols to the sum of inbound packets regardless of type and outbound ifoutucastpkts and ifoutnucastpkts plus 1 packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Percent Discards ifindiscards ifoutdiscards ifinucastpkts ifinnucastpkts ifindiscards ifinerrors inunknownprotos ifoutnucastpkts ifoutucastpkts 1 100 0 The ratio of the sum of inbound and outbound packets that are discarded to the sum of all inbound packets regardless of type and ifoutnucastpkts and ifoutucastpkts plus packet to prevent a potential divide by zero condition multiplied by 100 to yield a percentage Basic Metric Calculations 5 35 5 LAN and WAN ReportPacks Grade of Service Calculation Table 5 4 lists how the metrics are weighted in LAN Grade of Service GOS charts Table 5 4 LAN GOS Calculations GOS Calculation Utilization Errors Discards Weighting 60 20 20 Excellent Score Range 0 9 0 4 0 4 Good Score Range 10 19 5 6 5 6 Warning Score Range 20 39 7 9 7 9 Critical Score R
209. ncoln sanluis cotati Pkts sec 27538522 93 35379883 04 27480781 14 S7741747 27 11812231 66 33926123 71 27366425 47 4015947 52 Routers With Greatest Change In Throughput Thu Oct 21 1999 Change Rank Rank PrevRank Pct Discards Prev Pkts sec 27339710 09 35477323 17 27384531 32 37666101 63 11834926 91 33912132 44 27378710 53 4008552 08 1 ow OO A BP UN 4 6 2 2 5 4 1 4 7 7 3 3 6 5 8 8 3 73 2 48 3 73 1 86 2 58 3 73 3 48 2 52 The Routers with Highest Discard Rate table sorts by rank the interfaces by previous day s discard rate in ascending order Total throughput packets sec and bytes per second are also shown Device lincoln miami Sanjose sanluis houston cotati tulsa boston Pct Pkts Discarded 3 73 3 73 3 73 3 48 2 58 2 52 2 48 1 86 Routers with Highest Discard Rate Thu Oct 21 1999 Rank Prior Rank 1 1 2 3 4 5 6 8 ony aA AUN Pkts sec 33926123 741 27480781 14 27538522 93 27366425 47 11812231 66 4015947 52 35379888 04 37741747 27 Bytes sec 943368212 95 1182570806 32 1182195575 79 8534497 19 02 972846001 10 174640564 83 947004189 12 1213756965 47 The Routers with Greatest Change in Discard Rate sorts by change in rank the interfaces by the previous day s discard rate change in descending order Drastic increases in discard rate may suggest routers that should be watched while decreases document an effective remedy to a pr
210. nderstanding Remote Monitoring RMON RMON tracks network traffic that involves device other than the host device The monitoring can be local or remote Local Monitoring Most network devices can record events that affect them This is often done by an agent of some kind typically an SNMP agent An SNMP agent on a router can track the actions and performance on that router However such an agent cannot typically record what is occurring on other devices Additionally while the same agent can track how many packets were received and sent by the router it has no way to identify from which device the packets were received or to which device the packets were sent Thus local monitoring can tell you a great deal about the performance of a particular device but less about the health of a network segment Remote Monitoring Unlike a local agent an RMON agent is not typically concerned with the actions of its host device Instead an RMON device monitors a network segment recording all of the traffic it sees An RMON device is designed to track conversations between devices who sent what to whom and when and how those conversations took place This provides more definitive information on how segment bandwidth is used and can highlight potential trouble spots 9 2 Understanding Remote Monitoring RMON ReportPack Guide Report Descriptions The following sections describe each report in the RMON ReportPack in detail o D e
211. ng Daily Grade of Service Drill Down WAN 3 15 Capacity Planning Daily Utilization Drill Down WAN 3 16 Capacity Planning Daily Errors and Discards Drill Down WAN 3 17 Top Ten Busiest Bay Routers 0000 e eee ee 3 18 xi Figures xii 3 16 3 17 3 18 3 19 3 20 3 21 3 22 3 23 3 24 3 25 3 26 3 27 3 28 3 29 3 30 3 31 3 32 3 33 3 34 3 35 3 36 3 37 3 38 3 39 3 40 3 41 3 42 3 43 Top Ten Largest Bay Router Throughput Changes 3 19 Top Ten Bay Routers with Most Percentage of Discards 3 19 Top Ten Bay Routers with the Greatest Change in Percentage of DisGards am amaaa da E E Er E dae ald wae oe 3 20 QuickView Daily Grade of Service Frame Relay 3 22 Snapshot Snapshot Selection Table Frame Relay 3 23 Snapshot Grade of Service Frame Relay 3 23 QuickView Hourly Volume Frame Relay 3 24 QuickView Hourly Congestion 002 c eee eee eee 3 25 QuickView Distribution of Bandwidth Input 3 26 QuickView Distribution of Bandwidth Output 3 27 Near Real Time Router Selection List Bay Router 3 29 Near Real Time CPU Utilization Drill Down Bay Router 3 30 Near Real Time Memory Utilization Drill Down Bay Router 3 31 Near Real Time Buffer Utilization Drill Down Bay Router 3 32 Hot Spots Problem Summary for the Day Cisco
212. number of important service quality metrics appropriate to each technology Health scores for each metric are assigned and weighted according to their relative importance and added with other metric scores to calculate a final grade In the LAN Executive Summary Report 3 5 3 Analyzing TREND Reports ReportPack the key metrics contributing to overall health are Utilization Discards and Errors By assigning a grade and corresponding description for example Excellent Good Warning Critical to each hour of the day TREND provides an aggregate view of the stability or changes in overall health for the class Hourly Grade of Service All LAN Interfaces Mon Dec 20 1999 Critical Warning Good Excellent Mon 00 00 Mon03 00 Mon06 00 Mon09 00 Mon 12 00 Mon 15 00 Mon 18 00 Mon21 00 Tue 00 00 E Errors M Discards W Utilization Figure 3 4 Executive Summary Hourly Grade of Service Chart LAN Hourly Bandwidth Utilization Graph Figure 3 5 shows the Hourly Bandwidth Utilization line graph which shows executives volume expressed as a percentage of available bandwidth Hourly Bandwidth Utilization All LAN Interfaces Mon Dec 20 1999 22 80 17 10 11 40 Percent Utilization 5 70 0 00 00 06 00 12 00 18 00 00 00 Awg Percent Utilization Figure 3 5 Executive Summary Bandwidth Utilization Graph LAN 3 6 Executive Summary Report ReportPack Guide If the chart shows that current traffic is approaching
213. of availability changes Daily Hourly Hourly Availability wasp 1 Thu Oct 21 12 00 AM Fri Oct 22 12 00 Awl 100 00 75 00 60 00 25 00 0 00 00 06 00 12 00 18 00 00 00 Availability Service Level Figure 3 40 Service Level Hourly Availability Drill Down System Availability is a critical metric in most SLAs In principle availability is simply the percent of time that the system is up during any given time period Accessible data to measure availability and the quality of that data differ from device to device 3 44 Service Level Management Report ReportPack Guide Response Time Figure 3 41 shows the Response Time table which lists the systems with the highest response times Response Time Systems With Highest Respose Time Thu Oct 21 1999 Device Response Time leyte 100 00 hornet 100 00 laffey 100 00 nimitz 100 00 yorktown 100 00 eisenhower 100 00 enterprise 100 00 Figure 3 41 Service Level Response Time Drill Down System Response Time is another critical metric in most SLAs TREND processes each system s response time results and highlights problem systems with the worst or longest response times in this chart When vendor specific response time data from our partners is not available DeskTalk uses the time it takes for a device to respond to TREND s SNMP GET requests as a viable indicator of response time Because this measurement includes the desired device response time
214. ollowing types of installations New lInstalls a new version of the ReportPack for example System lt Version 4 0 gt Demo Installs a demonstration package with example report files for example System_Demo lt Version 4 0 gt Upgrade Installs an upgraded version of an existing ReportPack for example UPGRADE_System lt Version 4 05 For more information about the ReportPack options see Types of ReportPack Installations on page 1 5 7 Select the ReportPack option you want and click Add Installing ReportPacks 1 13 1 ReportPack Administration TREND begins the installation Ignore any No data found to load for table xxxx messages these are normal TREND will display a message when the installation is complete to ask if you want to begin to collect data The System package has been installed TREND will automatically configure data collection for this package No after midnight Would you like to start this process now Click No unless you have previously run IP Discovery For more information about IP Discovery see Chapter 5 of the TREND User s Guide Using the Property Import Feature The property import feature allows you to import user provided property informa tion that relates to the objects that a ReportPack reports on Specifically the property import feature allows you to set up the Executive Summary reports so that they only display data for devices belong
215. omponents that bring together the information needed to investigate potential network trouble spots pinpoint the source of the problem and determine possible resolutions Forecast Delivers a view into the future so that proactive corrections can be made to prevent service problems Provides a ranked list of all elements forecast to be within 90 days of reaching a threshold exception condition Reveals the num ber of days to threshold DTT and adds baseline utiliza tion values and 90 day projections for accurate forecast of budget requirements and capital equipment expenditure justification Each element is linked to multiple drill down components which show trends over the baseline period 1 of 2 Standard Reports ReportPack Guide Table 2 1 Standard Reports Report Description Capacity Plan Details the most overutilized and underutilized managed ning elements in the same report which indicates opportunities for load balancing to improve service levels without additional investment Enables the network manager to optimize existing resources and deliver improved service without increasing the budget Can be used to expose expensive underutilized resources that are targets for budget savings Drill down components examine element performance over time Top Ten Highlights the top volume contributors and the elements with the most severe problems Includes all key metrics for each of the top ten ma
216. omputer system in the summary table to view the following components of performance information for the selection The Grade of Service stacked bar chart shows weighted values for CPU memory and swap utilization and queue length for each day of the baseline period for the selected system See Table 12 7 for a description of the GOS weighting factors The Memory Utilization line graph shows the average memory utilization for the selected system for each day of the baseline period The CPU Utilization line graph shows the average CPU utilization for the selected system for each day of the baseline period The Swap Utilization line graph shows the average swap utilization for the selected system for each day of the baseline period The Total Swaps bar chart shows the number of swaps for the selected system for each day of the baseline period The Page Faults bar chart shows the number of page faults for the selected system for each day of the baseline period QuickView Report The QuickView report provides the network management staff with detailed infor mation on key metrics for individual systems with the highest GOS scores It iden tifies the ten systems with the highest worst grade of service scores Drilldowns show CPU utilization memory utilization swap utilization page faults and total swaps as well as number of processes running at the hourly level The QuickView report comprises the following components
217. on houston 3 Stock Room 290898951 4 4 23 79 boston 2 Northeast Sales 290036337 2 2 23 72 miami 2 Miktgs 240488729 3 6 21 84 sanjose 1 Sales 239835854 4 5 21 80 sanluis 2 Corporate Offices 239692493 5 4 21 79 miami 1 Sales 239041605 6 3 21 74 sanjose 2 ktg 239031279 7 7 19 57 houston 2 Accounting 167047995 12 13 13 65 tulsa 2 Tulsa Research Center 165739125 13 12 13 54 cotati 2 Cotati Sales 46660616 18 18 3 82 The LAN Interfaces with Greatest Change in Volume table shows the current day s volume previous day s volume and rankings for up to ten interfaces with the greatest changes in volume It ranks the interfaces by greatest change in volume An advantage to using this table is that it may highlight the newly installed devices and technologies that are approaching a logical threshold exceeding a threshold or impeding optimum utilization Report Descriptions 5 21 5 LAN and WAN ReportPacks 5 22 Device miami houston miami tulsa sanjose sanjose houston boston cotati sanluis The LAN Interfaces with Highest Grade of Service GOS Scores table ranks up to ten interfaces having the highest and therefore poorest GOS Interface LAN Interfaces With Greatest Change In Volume Description hlketgs Stock Room Sales Tulsa Research Center Sales hiktgs Accounting Northeast Sales Cotati Sales Corporate Offices Mon Dec 20 1999 Volume 240488729 290898951 23904
218. ons ReportPack Guide The Hourly Bandwidth Utilization line chart shows the percent of bandwidth utilization for each sample taken over the report period syoequoday o m gt z D a gt zZ The Hourly Grade of Service stacked bar chart shows GOS scores for the LAN metrics Utilization Errors and Discards or WAN metrics Input Utili zation Output Utilization Total Percent Errors and Total Percent Discards over the baseline period for the selected interface For more information about the GOS weighting calculations for LAN and WAN interfaces see Table 5 4 and Table 5 5 respectively Calculated values that are less than or equal to 1 show excellent performance Note For an in depth analysis of how to interpret the LAN or WAN Executive Summary report and for chart illustrations see Executive Summary Report on page 3 2 Forecast Report The Forecast report lists interfaces that are within 90 days of reaching a metric threshold condition indicating where near term correction can prevent service degradation For LAN interfaces the metric thresholds are as follows utilization 40 discard 5 or error 5 For WAN interfaces the metric thresholds are input and output utilization 80 The calculation for the threshold is from the baseline period Those interfaces closest to the 90 day threshold are listed first The forecast data is generated using the daily 95th percentile values for utiliz
219. ons Segment Dropped Events Segment Dropped Frames Report Descriptions 10 3 10 RMON2 ReportPack The Segment Errors line graph shows the total errors by type by hour for the interface selected from the Problem Summary for a Day table Segment Errors bluegrass ALL 3 Mon Dec 13 1999 0 01 Total 0 00 00 06 00 12 00 18 00 00 00 Hourly Scale Segment Giants Segment CRC Grors Segment Fragments Segment Jabbers Segment Runts The Frame Size stacked bar chart shows the total number of frames and the distribution of frame size by hour for the interface selected from the Problem Summary for a Day table Frame Size bluegrass ALL 3 Mon Dec 13 1999 95 082 00 63 388 00 31 694 00 Total Frames 0 Mon 00 00 Mon 09 00 Mon 18 00 Hourly Scale E 10241518 Bytes 512 1023 Bytes 256 511 Bytes 128 255 Bytes 65 127 Bytes i4 Bytes 10 4 Report Descriptions ReportPack Guide QuickView and Snapshot Reports The QuickView report provides detailed information on individual interfaces in a single report Specifically it lets you determine if your interface has possible probe resource problems due to high collision rates errors occurring on segments or high segment utilization The Snapshot report provides the same information as the Quick View report however it allows you to show detailed information about a particular interface The selected interface may be different than the interfaces
220. ons for the high change in volume Systems With Greatest Change In Volume Thu Oct 21 1999 Device Element Volume Previous Volume Volume Volume Rank Previous Rank Change Rank yorktowun 2 12195099049 85 12288955260 65 4 4 1 eisenhower a 9667909937 96 955340 1621 26 2 2 3 eisenhower 2 969146832 48 963566131 55 3 8 10 leyte 4 1137775468 31 1141841031 46 4 7 7 laffey 4 1221760234 53 1225594036 38 5 5 4 hornet 4 457219636 28 459759653 89 6 9 8 enterprise 1 407859790 15 409724021 40 7 44 44 yorktouun 1 1225213583 37 1223440369 62 8 4 5 wasp 1 457 197063 48 457569041 87 9 10 9 nimitz 2 1144631469 83 1194241107 29 10 6 6 i The Systems with Highest Grade of Service Scores summary table ranks the selected systems in ascending sequence by grade of service rank It shows the grade of service score grade of service rank previous rank daily averages for queue length and CPU memory and swap utilization for the top ten device and element pairs Systems With Highest Grade of Service GOS Scores Thu Oct 21 1999 Device Element GOSScore GOSRank PreviousRank CPU Utilization Memory Swap Queue Lengt Utilization Utilization nimitz 1 2 47 1 1 51 11 56 56 50 77 1 84 wasp 1 2 46 2 2 51 01 56 47 50 78 1 83 leyte 1 2 46 3 3 51 02 56 79 50 69 177 yorktown 4 2 25 4 4 34 21 58 63 50 48 1 54 latfey 4 2 24 5 5 34 09 58 51 50 47 1 59 hornet 1 1 27 6 6 11 66 27 69 22 32 0 69 eisenhower 41 1 18 F E 9 40 23 46 19 85 0 00 enterprise 4 1 17 8 8 9 46 23 28 19
221. onse Time line graph shows the daily response time for the selected Bay router for each sample for the preceding day Daily Hourly Hourly Response Time sanluis Unassigned Sun Dec 19 1999 1 484 50 7 1 187 60 90 70 milliseconds 593 80 4 296 90 0 __ _1 __ __14 __ 149 _ 4 __ __ 4 00 00 03 00 06 00 09 00 12 00 16 00 18 00 21 00 00 00 Data Source Source data for the Bay Router ReportPack is collected from the MIB I ifEntry table These are the same as those for the Router ReportPack See Table 7 1 for a list of variables Bayrtr resourcetotalentry table Table 7 1 lists the variables Table 7 1 Bay Router Variables Variable OID String wfResourceTotalSlot 1 3 6 1 4 1 18 3 3 2 5 7 1 1 The slot number is the index 1 of 2 7 26 Data Source ReportPack Guide Table 7 1 Bay Router Variables Variable OID String wfResourceTotalCpuUsed 1 3 6 1 4 1 18 3 3 2 5 7 1 2 The amount of CPU used by all the entities on this slot in centiseconds since the slot started Changed from INTEGER to Counter by DSI wfResourceTotalCpuldle 1 3 6 1 4 1 18 3 3 2 5 7 1 3 The amount of Idle CPU on this slot in centiseconds since the slot started Changed from INTEGER to Counter by DSI wfResourceTotalCpuMax 1 3 6 1 4 1 18 3 3 2 5 7 1 4 The maximum amount of CPU time possible on this slot in centiseconds since the slot
222. ory which would show the same statistics for all of the devic es in the network Figure 1 2 shows an example of an Executive Summary report that displays devices specifically for the Thunderbolt Inc customers Using the Property Import Feature 1 15 1 ReportPack Administration This report provides CFOs CIOs and other managers an overview AR ee oe Frame Relay of the performance of corporate Frame Relay networks Each chart T R E N D shows key metrics aggregated for all PVCs Key indicators of PVC Executive Summary performance are shown individually and combined into a Grade of Service to reveal PWC health at a glance Customer Acme P C Management Executive Summary Daily Volume for P Cs Thu Now 11 1999 Wed Dec 22 1999 97 00 77 60 58 20 Gigabytes 38 80 19 40 0 Nov 11 Nov 15 Nov19 Nowv23 Nov27 Dec01 DecO5 DecO9 Dec13 Dec17 Dec21 Figure 1 2 Devices Displayed for a Specific Customer The property import feature allows you to import CustomerID and CustomerName assignments for the objects being reported on The CustomerID assignments are used by The ReportPack when it processes data to provide aggregated summary information This information is used in Executive Summary reports TRENDWeb security functions to determine which data to provide to which users Using the Property Import Feature ReportPack Guide How Property Import Works Figure 1 3 shows a high level view of the ar
223. ose 1 20 4620000 512000 2 Acme 2 Unassigned RegionSupport atlanta 1 24 4620000 512000 2 Acme 2 Unassigned RegionMarketing atlanta 1 42 4620000 512000 2 Acme 2 Unassigned RegionMarketing lincoln 1 19 4620000 512000 2 Acme 2 Unassigned RegionFinance lincoln 1 79 4620000 512000 2 Acme 2 Unassigned RegionSupport Server orlando 1 30 4620000 512000 2 Acme 2 Unassigned RegionFinance orlando 1 50 4620000 10240002 Acme 2 Unassigned RegionFinance Server sanjose 1 40 4620000 512000 1 ABC 2 Unassigned RegionMarketing portland1 21 4620000 512000 1 ABC 2 Unassigned RegionAccounting torrancel 89 4620000 512000 1 ABC 2 Unassigned RegionSupport torrancel 90 4620000 512000 1 ABC 2 Unassigned RegionMarketing portlandl 4620000 512000 1 ABC F Unassigned RegionSales torrancel 88 4620000 512000 ABC Unassigned RegionSales 9 4 6 Figure 1 4 Sample Property File l dsi_target_name 2 dsi_table_key 3 port_speed Using the Property Import Feature 1 19 1 ReportPack Administration 4 pvc_cir 5 cust_id 6 customer_name 7 region_id 8 region_name 9 dsi_descr The property file has been populated for all columns except region_id and region_name and is also set up to display only devices belonging to Acme or ABC customers For a description of each property column see Table A 2 Displaying Data for Specific Customer Devices in Reports 1 20 To set up an Executive Summary report to show devices by customer
224. osition Defining the SourceDirectory The SourceDirectory directive tells TREND where to look for the property import files for a ReportPack It specifies the full path name to the directory and a file name specification that uses wildcard characters The wildcard characters allow TREND to match on file names The syntax of the SourceDirectory directive differs depending on the operating system you are using The TEEL file provides examples of the syntax for both Windows NT and UNIX To specify the SourceDirectory directive 1 Edit the ReportPack s TEEL file using a text editor like Notepad 2 Ensure that the SourceDirectory directive has the correct path and file name specification to match the incoming property import files Using the Property Import Feature ReportPack Guide Defining the SourceDisposition The SourceDisposition directive tells TREND what to do with the imported property file after processing it DeskTalk recommends that you use the move disposition to direct TREND to move the file to an archive directory D D U D Q Q gt 2 3 D 1 To specify the SourceDisposition directive 1 Edit the ReportPack s TEEL file using a text editor like Notepad 2 Ensure that the directory you specify for the SourceDisposition exists with write access Sample Property File Figure 1 4 shows a sample property file Frame_Relay_Property dat for the Frame Relay ReportPack Sanj
225. otal Exceptions Sanjose sanluis Sanjose sanluis 5 4 1 3 3 00 7 00 3 00 1 00 43 00 40 00 21 00 16 00 42 00 37 00 20 00 10 00 88 00 84 00 44 00 27 00 Double click on any Bay router listed in the table to update the following components with performance data for the selected router Report Descriptions ReportPack Guide The Hourly Grade of Service stacked bar chart shows weighted values for memory buffer and CPU utilization for the selected Bay router over each hour on the report day Table 7 3 gives the GOS weighting factors Hourly Grade of Service sanjose 3 Sun Dec 19 12 00 AM hion Dec 20 12 00 Aut Critical Warning Good Excelent Sun 00 00 Sun 09 00 Sun 18 00 E CPU Utilization E buffer Utilization Wimory Utilization The Bay Router Hourly CPU Utilization line graph shows the percentage of CPU utilization for each hour of the report day for the specified Bay router Bay Router Hourly CPU Utilization sanjose 3 Sun Dec 19 12 00 Ari hion Dec 20 12 00 Ar 95 60 71 70 47 80 Percent 23 90 0 00 00 06 00 12 00 18 00 00 00 Awg CPU Utilization P95 CPU Utilization Report Descriptions 7 19 yoed N ius ov lt D fo i ad es xe fo 7 7 Bay Router ReportPack The Bay Router Hourly Buffer Utilization line graph shows the average percentage of buffer utilization for each hour of the report day for the specified
226. p 1 Thu Oct 21 06 00 PM 4922 00 316 50 211 00 105 50 0 18 00 21 00 00 00 03 00 06 00 09 00 12 00 Time Scale Report Descriptions 12 15 _ N W lt 77 ca 3 D D O a U o 2 12 System ReportPack The CPU Utilization line graph shows the selected system s CPU utilization for each sample taken during the report period CPU Utilization wasp 1 Thu Oct 21 02 00 PM 83 20 62 40 41 60 20 80 16 00 18 00 21 00 00 00 03 00 06 00 09 00 Time Scale The Memory Utilization line graph shows the selected system s memory utilization for each sample taken during the report period Memory Utilization wasp 1 Thu Oct 21 06 00 PM 94 40 70 80 47 20 23 60 u 18 00 21 00 00 00 03 00 06 00 09 00 12 00 Time Scale 12 16 Report Descriptions ReportPack Guide The Page Faults line graph shows the selected system s page faults for each sample taken during the report period Page Faults wasp 1 Thu Oct 21 06 00 PM 143 409 20 107 656 90 71 704 60 36 852 30 0 18 00 00 00 06 00 12 00 Time Scale The Swap Utilization line graph shows the selected system s swap utilization for each sample taken during the report period Swap Utilization wasp 1 Thu Oct 21 06 00 PM 82 00 61 50 41 00 20 50 0 18 00 21 00 00 00 03 00 06 00 09 00 12 00 Time Scale Report Descriptions 12 17 n N w lt P 3 D O U D Q 12 System Report
227. packets for all devices divided by the total delta time for the aggregation period to yield packets per second Volume ifinoctets ifoutoctets inoctets and outoctets bytes for the selected managed elements are totaled for the specified aggregation period Basic Metric Calculations 6 27 6 Router ReportPack TREND 6 28 TREND 7 Bay Router ReportPack The Bay Router ReportPack reports statistics about CPU memory and buffer utilization and grade of service specifically for Bay Routers The Router ReportPack Chapter 6 includes Bay routers in its statistics for discards throughput and traffic volume The Bay Router ReportPack answers questions including Do the Bay routers provide an acceptable level of service Are performance problems due to high levels of CPU memory or buffer utilization or other causes yoed Which Bay routers are likely to degrade service in the near future if you do not take preventive action N w iy lt D e i po 2 od Report Descriptions The following sections describe each report in the Bay Router ReportPack in detail 7 1 7 Bay Router ReportPack Executive Summary Report 7 2 The Executive Summary report provides CFOs CIOs and other managers an overview of the performance of corporate Bay routers in the critical areas of volume and CPU memory and buffer utilization Each chart shows key metrics aggregated fo
228. perStatus 1 3 6 1 2 1 2 2 1 8 The current operational state of the interface The testing 3 state indicates that no operational packets can be passed ifLastChange 1 3 6 1 2 1 2 2 1 9 The value of sysUpTime at the time the interface entered its current operational state If the current state was entered prior to the last reinitialization of the local network management subsystem then this object contains a zero value ifInOctets 1 3 6 1 2 1 2 2 1 10 The total number of octets received on the interface including framing characters ifInDiscards 1 3 6 1 2 1 2 2 1 13 The number of inbound packets which were chosen to be discarded even though no errors had been detected to prevent their being deliverable to a higher layer protocol One possible reason for discarding such a packet could be to free up buffer space ifInErrors 1 3 6 1 2 1 2 2 1 14 2 of 3 6 22 Data Source ReportPack Guide Table 6 1 Variables Polled from MIB II ifEntry Table Variable OID String The number of inbound packets that contained errors preventing them from being deliverable to a higher layer protocol ifOutOctets 1 3 6 1 2 1 2 2 1 16 The total number of octets transmitted out of the interface including framing characters 5 fe ifOutDiscards 1 3 6 1 2 1 2 2 1 19 The number of outbound packets which were chosen to be discarded even D Sree
229. plus the processing time of the device s SNMP agent plus the round trip network delay you can gain more insight by reviewing response time trends or changes rather than focusing on the raw response time value The Service Level Management report assists in this by presenting daily response time over the baseline period see Figure 3 42 for long term trend analysis and hourly response time over the past day see Figure 3 42 for detailed investigation so you can see the relative changes in response time Double click any system in the Response Time table to reveal the detailed daily and hourly response time drill down charts swioday wo gt lt S gt Ko J m z s Service Level Management Report 3 45 3 Analyzing TREND Reports Daily Response Time Drill Down Figure 3 42 shows the Daily Response Time line graph which shows the network response time over the baseline period for the selected system Daily l Hourly Daily Response Time wasp 1 Fri Sep 10 1999 Thu Oct 21 1999 125 20 93 90 62 60 milliseconds 31 30 o Sep10 Sep17 Sep24 Oct01 Oog Oct 15 Figure 3 42 Service Level Daily Response Time Drill Down System 3 46 Service Level Management Report ReportPack Guide Hourly Response Time Drill Down Figure 3 43 shows the Hourly Response Time chart which shows the network response time for the selected system for each hour yesterday bay Hourly Hourly R
230. ption Volume Amount of information that flows through a managed element For example for Frame Relay volume is the total inoctets outoctets in Gigabytes for all Permanent Virtual Circuits PVCs Utilization The amount of information actually flowing through a managed element divided by the capacity of the device Utilization is given in terms of bandwidth for Frame Relay PVCs and in terms of percentage of CPU time memory and buffer space for LAN or WAN interfaces Throughput Amount of information that passes through a managed element completely and correctly over time The metric is given in millions of bits per second Mbps 1 of 3 2 6 Metric Classes ReportPack Guide Table 2 2 Metric Classes Metric Description Congestion Measures the number of frames that experi ence congested areas on the network in both forward and backward directions Indicates the level of impediments to throughput Grade of Service GOS Weighted scores for selected metrics that together add up to a number in one of the following ranges that signifies a GOS score The following is an example of GOS grading based on a scale of four points N O lt i Range GOS Score 0 0 1 0 Excellent 1 1 2 0 Good 2 1 3 0 Warning 3 1 4 0 Critical Latency Amount of time it takes for information to pass through a managed element or the network Availability
231. r a description of GOS weighting factors The Bay Router Daily CPU Utilization line graph shows the average and 95th percentile CPU utilization over each day of the baseline period for the selected router The Bay Router Daily Buffer Utilization line graph shows the average and 95th percentile buffer utilization over each day of the baseline period for the selected router yoed N ius o lt D e 7 Po i xe e Di The Bay Router Daily Memory Utilization line graph shows the average and 95th percentile memory utilization over each day of the baseline period for the selected router Report Descriptions 7 17 7 Bay Router ReportPack Hot Spots Report 7 18 The Hot Spots report provides the network manager and other members of the network management staff a list of Bay routers that have exceeded threshold conditions The Hot Spots report comprises the following charts The Bay Router Problem Summary summary table displays the top ten Bay routers in descending order by total number of exceptions CPU memory buffer utilization exceptions for the report day An exception occurs when an interface exceeds its defined CPU 85 memory 80 or buffer 80 utilization threshold in any sample taken during the report day Router Exception Bay Router Problem Summary Sun Dec 19 1999 Sun Dec 19 1999 Slot CPU Utilization Memory Utilization Buffer Utilization Exceptions Exceptions T
232. r all Bay routers Key indicators of performance are shown individually and combined into a Grade of Service chart to reveal Bay router health at a glance The Executive Summary report comprises the following components The Bay Router Inventory summary table shows the number of LAN WAN and Other interfaces total interfaces LAN WAN Other number of Bay routers that are managed total volume for all Bay router LAN WAN and Other interfaces Bay Router Inventory Sat Jan 01 2000 Sat Jan 01 2000 Vendor WAH Interfaces LAN Interfaces Other Total Routers Managed Volume in Interfaces Interfaces Terabytes Bay 1 2 0 3 1 26 43 The Hourly Volume bar chart shows the total volume for all Bay routers for each hour yesterday Hourly Baseine Daily Hourly Volume All Bay Routers Sat Jan 01 12 00 A Sat Jan 01 11 00 PM 1 80 iy Gigabytes bad a oa 0 00 00 06 00 12 00 18 00 Dwolume in Gigabytes Report Descriptions ReportPack Guide The Hourly Baseline line chart shows the average volume for each hour of the day over the baseline period You can compare total volume for each hour of the report day with the baseline for that hour for the report period Hourly Baseline Daily Hourly Baseline All Bay Routers Sun Now 21 1999 Sun Now 21 1999 1 80 1 20 0 60 0 00 00 06 00 12 00 18 00 Daily Scale The Daily Volume bar chart shows the total volume for all interfaces on
233. r graph shows the interfaces with the lowest availability for the report day Availability is essentially computed as the ratio of the time that an interface is functional to the total time the interface is up displayed as a percentage See Daily Availability Drill Down on page 3 42 for a discussion of the factors that are used to compute availability Report Descriptions 5 23 syoequoday o gt Z 2 gt a 5 LAN and WAN ReportPacks Availability LAN Interfaces With Lowest Availability For The Day Mon Dee 20 1999 Device Interface Availability miami 2 100 00 Sanjose 41 100 00 sanjose 2 100 00 cotati 2 100 00 houston 2 100 00 houston 3 100 00 boston 2 100 00 tulsa 2 100 00 sanluis 2 100 00 Double click on a interface in this chart to update the following chart with information for the selected interface The Daily Availability line graph shows the average daily availability for the selected interface for each day of the report period Daily Hourly Availability miami Sales Tue Now 09 1999 hon Dec 20 1999 100 00 80 00 60 00 Percent 40 00 20 00 0 Nov 09 Nov 23 DecO7 Availability 5 24 Report Descriptions ReportPack Guide The Hourly Availability line graph shows the average hourly availability of the selected interface for each hour for each sample taken during the report period syoequoday o m gt Z 5 2 gt a Dail
234. rage CPU Utilization The Hourly Buffer Utilization line graph shows the 95th percentile of the 95th percentile of all Buffer Utilization samples collected for each hour of the report day for each buffer size There are five buffer sizes huge large big medium and small Hourly Buffer Utilization All Cisco Routers Sun Dec 19 1999 80 00 40 00 Percent 0 00 00 06 00 12 00 18 00 Hourly Scale Huge Buffer Utilization Large Buffer Utilization Big Buffer Utilization WNMedium Buffer Utilization Small Buffer Utilization D S e U o i x J9 NOY ODSID 8 Report Descriptions 8 5 8 Cisco Router ReportPack The Hourly Exception Counts bar chart shows the number of exceptions for CPU Utilization Buffer Utilization Physical Memory Utilization and Buffer Errors for all Cisco routers for each hour of the report day Hourly Exception Counts All Cisco Routers Sun Dec 19 1999 10 20 6 80 3 40 0 00 00 02 00 04 00 06 00 08 00 10 00 12 00 14 00 16 00 18 00 20 00 22 00 E Buffer Utilization E Physical Memory Utilizaiton W CPU Utilization Wi Buffer Errors Forecast Report 8 6 The Forecast report lists all Cisco routers that are within 90 days of reaching a utilization threshold Those Cisco routers closest to the 90 day threshold are listed first Each managed element is hot linked to multiple graphs for analysis over the baseline period The default baseline period is the s
235. rank Volume and health change charts alert users to the PVC s greatest changes in volume suggesting potential capacity overloading if increasing trends continue The volume change report also may highlight which newly installed devices and technologies are approaching a logical threshold and which may be exceeding a threshold or which are underutilized Drastic increases in grade of service scores suggest additional PVCs to be watched while decreases document an effective remedy to a previously negative condition The Top Ten Report comprises the following The PVCs with Highest Volume table shows up to ten PVCs sorted by the current rank in ascending based on the report day s volume PVCs With Highest Volume Wed Dec 22 1999 Device Interface DLCI Description Volume Volume Rank Previous Rank Utilization Congestion orlando 1 50 Accounting Sewer 6587821902 63 4 1 469 77 0 00 torance 1 89 Support 156539485 82 2 3 250 22 0 02 portland 1 21 Accounting 156497759 61 3 4 249 83 0 02 portland 1 30 Sales 1656121302 16 4 2 249 41 0 02 torrance 1 90 Marketing 87768136 22 5 8 140 13 0 04 Sanjose 1 20 Support 87762855 55 6 5 140 21 0 04 sanjose 1 40 Marketing 87698904 20 7 7 140 04 0 04 torrance 1 88 Sales 87585468 15 8 6 140 08 0 04 orlando 1 30 Accounting 60177092 68 9 9 96 14 0 08 lincoln 1 79 Support Server 35373290 12 10 11 56 52 0 77 4 18 Report Descriptions ReportPack Guide The PVCs with Greatest Change in Volume table shows
236. rank for the report day and the previous day Previous Rank is also given along with the report day s percentage of discards The throughput is shown in packets per second and bytes per second Cisco Routers With Highest Throughput Sun Dec 19 1999 Router Pkts sec Rank Prev Rank Bytes sec Pct Discards boston 37741747 27 1 1 1213756965 47 1 86 tulsa 36379888 04 2 2 947004189 12 2 48 lincoln 33926123 71 3 3 4943368212 95 3 73 miami 27480781 14 5 4 1182570806 32 3 73 houston 11812231 66 7 7 972846001 10 2 58 cotati 4015947 52 8 8 174640564 83 2 52 The Cisco Routers with Greatest Change in Throughput table shows the Cisco routers with the largest change in rank from the preceding day to the report day The Cisco router with the largest change in rank is listed first Cisco Routers With Greatest Change In Throughput Sun Dec 19 1999 Router Pkts sec Prev Pkts sec Change Rank Rank PrevRank Pct Discards tulsa 35379883 04 36477323 17 2 2 2 2 48 miami 27480781 14 27384631 32 3 5 4 3 73 boston 37741747 27 37666101 63 4 1 1 1 86 houston 11812231 66 11834926 91 5 7 7 2 58 lincoln 33926123 71 33912132 44 6 3 3 3 73 cotati 4015947 52 49008552 08 3 8 8 2 52 D o S e U o i x J9 NOY ODSID 8 Report Descriptions 8 25 8 Cisco Router ReportPack The Cisco Routers with Highest Discard Percentage table ranks the ten Cisco routers with the largest average percentage of discard scores for the report day
237. rds boston Acme s Personnel Office Thu Oct 21 1999 1 130 199 00 904 159 20 678 119 40 452 079 60 Total Discards 226 039 80 0 Thu 00 00 Thu 09 00 Thu 18 00 The Percent Discarded Packets line graph shows the percentage of discarded packets over each hour of the report day for the selected router Percent Discarded Packets boston Acme s Personnel Office Thu Oct 21 1999 2 40 1 80 3 m amp 1 20 a 0 60 0 00 00 06 00 12 00 18 00 00 00 6 10 Report Descriptions ReportPack Guide Hot Spots Report The Hot Spots report displays the top ten routers in descending order by the total number of exceptions that exceed predetermined threshold values for detailed investigation of potential network trouble spots Types of exceptions that contribute to the total exception count include utilization errors and discards exceptions The Hot Spots report comprises the following components The Problem Summary for the Day table shows the routers with the highest exceptions counts ranked by total number of utilization discard and error exceptions Problem Summary for the Day Routers With Highest Discards Errors Thu Oct 21 1999 Router Vendor Avg Percent TotalDiscards Avg Percent Total Errors Discards Errors a a73 azma amo aaow sanjose Bay 3 73 32811359 3 86 33793545 lincoln Cisco 3 73 34531483 3 62 33417331 sanluis Bay 3 48 28781990 3 50 27657464 houston Cisco 2 58 7884068 1 21 3081164 cotati Cisco 2 52 3543
238. reates all the data tables key tables and stored procedures Adds lines to the t rend_timer sched file which schedules the ReportPack s t rend_proc files to run Installs the TRENDweb reports To run Autopilot 1 Ensure that you are logged in as the trendadnm user 1 10 Installing ReportPacks ReportPack Guide 2 Click the Start button and then point to Programs Point to the folder that contains TREND and then click TREND The TREND window is displayed D U D 2 gt 2 3 2 D 1 3 Click Autopilot The Autopilot dialog box is displayed Components REMOVE Installing ReportPacks 1 11 1 ReportPack Administration 4 Click Add The TREND Report Packs dialog box is displayed F TREND Report Packs _ oy x C Mtrend packages 1 12 Installing ReportPacks ReportPack Guide 5 Double click the name of the ReportPack for example System in the Directory list The TREND ReportPacks dialog box displays the available ReportPack s D o U D Q 2 gt 2 3 D 1 wha ix CMtrend packages System System ap System_Demo ap UPGRADE_System ap Components System lt Version 4 0 gt System_Demo lt Version 4 0 gt UPGRADE_System lt Version 4 0 gt 6 Click the name of the ReportPack you want to install for example System lt Version 4 0 gt in the ReportPack s list You can choose one of the f
239. recast database statistics generated from TRENDit and TRENDsum TRENDit populates the rate data table Rsi_dbstats from the data inserted into the raw table dsi_dbstats by the TREND database client applications TRENDsum populates the SDdsi_database_util and SD42SDdsi_database_util_forecast tables that contain the daily and forecast data Source data for the TREND Database ReportPack is collected from the following tables Data Source ReportPack Guide SDdsi_database_util table SD42SDdsi_database_util_forecast table SDdsi_database_util Table For the SDdsi_database_util table TRENDsum calculates the daily average and maximum database size database used percentage of database used transaction log size transaction log used and percentage of transaction log used Table 11 1 lists the variables Table 11 1 TREND Database Daily Average and Maximum Variables Variable Description AV Gdbsizekb The average size of the database in kilobytes over each day MA Xdbsizekb The maximum size of the database in kilobytes over each day AV Gpct_dbused The average percentage of the database that is used over each day MAXpct_dbused The maximum percentage of the database that is used over each day AVGkb_dbused The average number of kilobytes the database used over each day MAXkb_dbused The maximum number of kilobytes the database used over each day AVGlogsizekb The
240. report comprises the following components The Estimated Days to Threshold DTT summary table lists the filesystems that are within 90 days of 100 percent utilization It shows the days to threshold Report Descriptions ReportPack Guide for filesystem utilization current filesystem utilization and projected utilization in 90 days for each filesystem Estimated Days to Threshold DTT Filesystems Within 90 Days Of 90 Utilization Device Filesystem DTT Filesystem Current Filesystem Projected Util Util Utilization in 90 Days yorwtown exportfhome 0 00 95 00 95 00 Double click on a filesystem in the summary table to view the following components of performance information for the selection The Grade of Service stacked bar chart shows weighted values for queue length and CPU memory and swap utilization for each day of the baseline period for the selected filesystem See Table 12 7 for a description of the GOS weighting factors Grade Of Service GOS wasp C Fri Sep 10 1999 Thu Oct 21 1999 Excellent Sep 09 Sep 23 Oct OF M Queue Length E swap utilization memory Utilization E CPU Utilization Report Descriptions 12 5 _ N 2 lt P 3 D O U D Q 12 System ReportPack The Filesystem Utilization line graph shows the average filesystem utilization for the selected filesystem for each day of the baseline period Filesystem Utilization wasp C Fri Sep 10 19
241. reports on the performance of individual Cisco routers Collectively the reports help you identify problems on your Cisco routers which include resource failures not enough memory or a router that is unable to handle the amount of traffic being generated The reports can help you troubleshoot problems by showing what types of errors are occurring on an interface queue drops utilization and protocol distribution The Cisco Router ReportPack answers questions including Do the Cisco router networks provide an acceptable level of service to customers or end users Are there overloaded Cisco routers If so are there underutilized Cisco routers that can help balance the load Are performance problems due to high levels of discards errors excessive utilization or other causes What are the top ten Cisco routers contributing to the poor health of the network 8 1 D o me fe U o P x JOINOY ODSID 8 8 Cisco Router ReportPack Which Cisco routers are likely to degrade service unless you take preventive action Each report consists of a combination of charts graphs and tables To produce the reports the Cisco Router ReportPack discovers all Cisco routers collects their volume and discard data and then processes the data automatically Report Descriptions The following sections describe each report in the Cisco Router ReportPack in detail Executive Summary Report The Executive Sum
242. rises the following components The Snapshot summary table shows the time period device CPU utilization memory utilization total page faults and swaps Double click on a system to update the following components with performance data for the selection Report Descriptions 12 13 _ N w lt P 3 D O U D Q 12 System ReportPack The Number of Processes line graph shows the number of processes for each sample period of the report day for the selected system The CPU Utilization line graph shows the CPU utilization for the selected system The Memory Utilization line graph shows the memory utilization for the selected system The Page Faults bar chart shows the number of page faults for the selected system over each hour of the report day The Swap Utilization line graph shows the swap utilization for the selected system The Total Swaps bar chart shows the number of swaps for the selected system over each hour of the report day Near Real Time QuickView and Snapshot Reports 12 14 The Near Real Time Quick View report provides up to date system performance statistics up to the last SNMP poll The report does not rely on nightly summaries therefore it provides instant reporting on collected data Network managers and analysts can go to this report to select a system from a table and assess its status in the following performance areas CPU utilization number of processe
243. rly Grade of Service Drill Down 0000000 3 34 CPU Utilization Drill Down 0 00 000 eee eee 3 36 Buffer Utilization Drill Down 00 0 c eee eee eee 3 37 Memory Utilization Drill Down 0 00000 3 38 Percent Discards Drill Down aaua c eee eee 3 39 Buffer Faults Drill Down 0 0000 aeaaee 3 39 Exception Detail Drill Down 0 0000 e eee eee 3 40 Service Level Management Report 000 eee 3 41 Availability Table a n 0 eee eee 3 42 Daily Availability Drill Down 0 0000 e eee eee 3 42 Hourly Availability Drill Down 0 000000 eee eee eee 3 43 Response Time 0c eee eee 3 45 Daily Response Time Drill Down 0 0000s e eee 3 46 Hourly Response Time Drill Down 0000 0 eae 3 47 Contents vi Frame Relay ReportPack 0002 ee eee eee eee neces 4 1 Report Descriptions 0 0 00 c eee tees 4 2 Executive Summary Report 0 0 eee eee ees 4 2 Forecast Repor teerien E a ain eaten Rete ee Pn ne 4 6 QuickView and Snapshot Reports 0 000000 ee eae 4 7 Near Real Time QuickView and Snapshot Reports 4 9 Capacity Planning Report 0 00 e eee eee 4 12 Hot Spots Report eee ccccce ce eee Rhee i die dole ee 4 15 Top Ten Report a eats ected Peeled la wed ened bea dda 4 18 Service Level Management Report 00 ee 4 20 Data
244. router Bay Router Hourly Buffer Utilization sanjose 3 Sun Dec 19 12 00 AM Mon Dec 20 12 00 AM 93 80 74 10 Percent 24 70 0 00 00 06 00 12 00 18 00 00 00 Awg Buffer Utilization P95 Buffer Utilization The Bay Router Memory Utilization line graph shows the average percentage of memory utilization for each hour of the report day for the specified router Bay Router Memory Utilization sanjose 3 Sun Dec 19 12 00 Ari Mon Dec 20 12 00 A 100 00 75 00 50 00 Percent 25 00 0 00 00 06 00 12 00 18 00 00 00 Avg Memory Utilization P95 Memory Utilization 7 20 Report Descriptions ReportPack Guide The Exception Detail table shows the time the sample was taken for each CPU buffer and memory utilization exception and the values for the offending conditions Table 7 2 describes how each metric is calculated Exception Detail Report sanjose 3 Sun Dee 19 12 00 AM Mon Dec 20 12 00 AM Time Period Router Slot Description CPU Utilization Memory Buffer Utilization Utilization Sun Dec 19 12 00 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 12 15 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 12 30 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 12 45 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 01 00 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 01 15 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 01 30 AM sanjose 3 82 21 100 00 98 75 Sun Dec 19 01 45 AM sanjose 3 82 21 100 00 98 75 Sun Dec 1
245. rs 134 70 65 84 134 70 32 24 134 70 150 17 134 70 32 5 134 70 32 5 134 70 32 5 134 70 32 5 134 70 65 206 134 70 65 106 209 219 153 173 10 10 Mon Dec 13 1999 Source Address Target Address Sent Mbytes 134 70 32 251 134 70 150 17 134 70 32 24 209 219 153 173 134 70 65 206 134 70 180 100 134 70 65 106 134 70 32 5 134 70 32 5 134 70 32 5 45 13 25 08 15 32 14 91 14 50 14 50 14 42 14 18 14 13 13 94 Report Descriptions ReportPack Guide The Top Ten Talkers table lists the IP address and the total megabytes sent and received for the addresses that sent the most data to all destinations on the preceding day Top Ten Talkers Mon Dec 13 1999 Src Address Probe IFNumber WNetworkLayer Sent Mbytes Rev Mbytes 134 70 32 5 bluegrass 3 ip 1 72 49 72 02 134 70 65 84 bluegrass 3 ip 1 43 48 3 01 hire 134 70 150 17 bluegrass 3 ip 1 26 09 29 32 134 70 32 24 bluegrass 3 ip 1 25 44 17 43 2 134 70 65 106 bluegrass 3 ip 1 16 01 15 85 e 209 219 153 173 bluegrass 3 ip 1 15 47 15 91 5 134 70 180 100 bluegrass 3 ip 1 15 02 15 46 py 134 70 65 206 bluegrass 3 ip 1 14 89 15 26 iS 134 70 180 113 bluegrass 3 ip 1 13 93 14 32 134 70 32 1 bluegrass 3 ip 1 8 77 7 66 x The Top Ten Listeners table lists the IP address and the total megabytes sent and received for the addresses that received the most data from all destinations on the preceding day Top Ten Listeners Mon Dec 13 1999 Sre Address Probe IFNumb
246. rsion 4 0 gt LAN_WAN_Connectivity_Demo lt Version 4 0 gt Bay_Router_Demo lt Version 4 0 gt RMON_Demo lt Version 4 0 gt RMON2_Demo lt Version 4 0 gt Router Demo lt Version 4 0 gt Select the ReportPack for example System lt Version 4 0 gt and click Remove The Remove ReportPack dialog box is displayed Remove Report Pack j server reporting pro Collector Definitions empire_cpuStatsEntry empire_diskStatsEntry empire_devTableEntry empire_kernelperf 4 Click Full Removal and click OK This deletes all the data tables and modifies trendt imer sched as appropriate 1 26 Removing the ReportPack ReportPack Guide Selecting Deferred Targets or Elements for a Snapshot Report D U D Q x gt 2 3 D 1 A snapshot report is one of the eight standard reports in a ReportPack It allows you to view statistics for a selected target or element To select a target or element for a Snapshot report 1 Select the Snapshot report from the Report Listing area of the TRENDview Pro window The Choose one or more Targets or Elements dialog box is displayed 4 Choose one or more Targets or Elements olx Select Targets Elements for cisco_router_snapshot qss Available Targets Elements Selected T argets Elements EHC atlanta 2 boston 2 cotati 2 houston D lincoln I newportnews 2 newyork 2 orlando 2 portland EH raleigh EHC s
247. rt 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 26 Support 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 18 Marketing 0 00 0 00 0 00 0 00 COPBKR_BCIO1 202104 0 28 Accounting 0 00 0 00 0 00 0 00 Double click on a device in either table to update the content of the following drill down components with performance data for the device The Grade of Service bar chart shows Grade of Service GOS scores for the selected PVC Table 4 3 describes how Grade of Service is calculated Grade of Service GOS orlando Accounting Server Thu Nov 11 1999 Wed Dec 22 1999 Critical Warning E Congestion E out utilization Win Utilization Excellent Nov11 Now15 Nov19 Nov23 Nov2 DecO1 DecO5 DecO9 Dec13 Dec17 Deco21 gt TI 3 D U 2D O0 x lt D Report Descriptions 4 13 Z 1 4 Frame Relay ReportPack The Utilization line graph shows In Utilization and Out Utilization calculated using CIR and ifSpeed for the selected PVC for each day of the baseline period Utilization orlando Accounting Server Thu Nov 11 1999 Wed Dec 22 1999 375 60 250 40 Percent 125 20 Nov 11 Now 25 DecO9 P95 In Util CIR P95 Out Util CIR P95 In Util Ifspeed P95 Out Util Ifspeed The Congestion line graph shows congestion for the selected PVC for each day of the baseline period Congestion orlando Accounting Server Thu Nov 11 1999 Wed Dec 22 1999 2 472 80 1 854 60 1 236 40 61
248. rt day The Congestion line graph shows congestion for the selected PVC for each sample taken on the report day QuickView and Snapshot Reports The Quick View report identifies the top ten PVCs with the highest worst grade of service scores according to Grade of Service score Network managers and analysts can go to this report to select a PVC from a grade of service chart and assess its status in several performance areas Drill down charts show daily distribution of bandwidth as polled volume against the baseline and congestion at the as polled level Alternatively the Snapshot report shows the same information however you select PVCs you want to view from a pick list when you invoke the report For more information about selecting a PVC for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 U o P x Report Descriptions 4 7 Woday Aejay ewei4 v 4 Frame Relay ReportPack 4 8 Note For an in depth analysis of how to interpret the Frame Relay Quick View and Snapshot reports and for chart illustrations see QuickView and Snapshot Reports on page 3 20 The QuickView and Snapshot reports comprise the following charts QuickView The Daily Grade of Service stacked bar chart shows the PVCs with the highest daily GOS scores Table 4 3 describes how Grade of Service is calculated Snapshot The Snapshot Selection table lists the time period
249. rt day for the selected interface 5 26 Report Descriptions ReportPack Guide Daily Hourly Response Time miami Sales Mon Dec 20 12 00 A milliseconds a t 24 80 0 00 00 06 00 12 00 18 00 00 00 Data Source Source data for the LAN and WAN ReportPacks is collected from the MIB H ifEntry table Table 5 1 lists the variables Table 5 1 LAN and WAN Connectivity Variables Variable OID String ifIndex 1 3 6 1 2 1 2 2 1 1 A unique value for each interface Its value ranges between 1 and the value of ifNumber The value for each interface must remain constant at least from one re initialization of the entity s network management system to the next re initialization 1 of 5 Data Source 5 27 syoequodey o m gt Z 2 gt Z 5 LAN and WAN ReportPacks 5 28 Table 5 1 LAN and WAN Connectivity Variables Variable OID String ifDescr 1 3 6 1 2 1 2 2 1 2 A textual string containing information about the interface This string should include the name of the manufacturer the product name and the version of the hardware interface ifType 1 3 6 1 2 1 2 2 1 3 The type of interface distinguished according to the physical link protocol s immediately below the network layer in the protocol stack ifMtu 1 3 6 1 2 1 2 2 1 4 The size of the largest datagram which can be sent received on the interface speci
250. rtPack directory may contain a folder for the ReportPack its demo ver sion and th e ReportPack s upgrade Figure 1 1 shows an example of what the directory structure for the System ReportPack might look like TREND DPIPE_HOME Figure 1 1 packages System System ap System_Demo ap UPGRADE_System ap System ReportPack Example Directory Structure Installing ReportPacks ReportPack Guide Types of ReportPack Installations You can perform the following types of ReportPack installations new upgrade or demo D U D Q 2 gt 2 3 D 1 New Installation A new installation installs a new version of the ReportPack For more information about installing a new ReportPack see Running Autopilot on page 1 10 Note Before you install a ReportPack ensure that no previous versions of it are installed If there is a previous version remove it using Autopilot or you can install an upgrade version of the ReportPack if one is available Upgrade Installation An upgrade installation upgrades an existing version of a ReportPack You can install an upgrade version of the ReportPack for example UPGRADE_System lt Version 4 0 gt if you have a V3 5 version of the ReportPack For more information about installing an upgrade see Running Autopilot on page 1 10 Installing ReportPacks 1 5 1 ReportPack Administration The upgrade installation checks for
251. rty File Formats 000 cee ee eee eee eee eee A 1 LAN WAN Connectivity ReportPack 000 eee eee A 2 Frame Relay ReportPack aaan eee eee eee ee A 3 Router ReportPack 0 0 00 eee A 4 INGOX Ae Stee a Serie thee SE Cae ie GS ERA Index 1 TREND F TREND gures System ReportPack Example Directory Structure 1 4 Devices Displayed for a Specific Customer 1 16 Architecture of the Property Import Feature 1 17 Sample Property File 0 0 0 0 ee 1 19 TREND Report Layout 00 00 2 2 Forecasting Least Squares Linear Regression 2 9 Sample Chart for Frequency Distribution 2 15 Executive Summary Daily Volume Chart LAN 3 3 Executive Summary Hourly Volume Chart LAN 3 4 Executive Summary Hourly Baseline Chart LAN 3 5 Executive Summary Hourly Grade of Service Chart LAN 3 6 Executive Summary Bandwiath Utilization Graph LAN 3 6 Forecast Estimated Days to Threshold Table Frame Relay 3 8 Forecasts Based on 95th Percentile vs Average Utilization 3 9 Forecast Grade of Service Drill Down Frame Relay 3 10 Forecast Utilization Drill Down Frame Relay 3 11 Forecast Congestion Drill Down Frame Relay 3 12 Capacity Planning Overutilized WAN Interfaces Summary 3 14 Capacity Planni
252. s 100 Frames with Forward Explicit Congestion Notification fecn plus Frames with Backward Explicit Congestion Notification becn divided by the number of outframes plus inframes times 100 to yield a percentage fecns becns inframes and outframes are collected as a number of frames since the virtual circuit was created The number of frames with FECNs and BECNs respectively is collected from the frCircuitReceivedFECNs and frCircuitReceivedBECNs objects respectively in the Frame Relay MIB Volume total inoctets outoctets inoctets plus outoctets for the selected PVCs are totaled for the specified aggregation period 2 of 2 Basic Metric Calculations 4 27 yed A TI o 3 D D o o lt D me fe T 4 Frame Relay ReportPack Grade of Service Calculation Table 4 3 lists how the metrics are weighted in Frame Relay Grade of Service GOS charts Table 4 3 Frame Relay GOS Calculations Input Output Input Output GOS Utilization Utilization Utilization Utilization Calculation CIR CIR Congestion ifSpeed ifSpeed Weighting 20 20 60 20 20 Excellent 0 80 0 80 50 0 40 0 40 Score Range Good 80 100 80 100 50 1 40 60 40 60 Score Range Warning 100 100 150 1 2 60 80 60 80 Score 150 Range Critical Over Over Over 2 80 100 80 100 Score 150 150 Range For more information about
253. s Discards ifindiscards The number of inbound plus ifoutdiscards outbound packets that are discarded for reasons other than containing errors The source of this metric is the data collected for the LAN WAN ReportPacks 2 of 2 Basic Metric Calculations 7 29 yoed N w iy lt D e i po iv 2 od 7 Bay Router ReportPack Grade of Service Calculation Table 7 3 lists how the metrics are weighted in Bay router Grade of Service GOS charts Table 7 3 Bay Router GOS Calculations CPU Memory Buffer GOS Calculation Utilization Utilization Utilization GOS weight factor 20 40 40 GOS score 1 Excellent 0 84 0 20 0 20 GOS score 2 Good 85 89 21 60 21 60 GOS score 4 Warning 90 95 61 80 61 80 GOS score 5 Critical 95 100 81 100 81 100 For more information about the Grade of Service metric see Table 2 2 on page 2 6 For more information about Grade of Service as it applies to each type of report see Analyzing TREND Reports on page 3 1 7 30 Grade of Service Calculation TREND Cisco Router ReportPack The Cisco Router ReportPack provides comprehensive data collection and reporting to support management of Cisco router networks It includes compound reports for capacity planning service level reporting forecasting and troubleshooting It also provides an executive summary on the performance of all Cisco routers combined and detailed
254. s memory utilization page faults swap utilization and filesystem utilization Alternatively the Snapshot report shows the same information however you select systems you want to view from a pick list when you invoke the report For more information about selecting a system for a Snapshot report see Selecting Deferred Targets or Elements for a Snapshot Report on page 1 27 The QuickView and Snapshot reports comprise the following charts The System Selection List table shows each system and its performance statistics based on the data polled for the last 6 hours The statistic shown for Report Descriptions ReportPack Guide each metric is the average of all the data collected for that statistic over the 6 hour period System Selection List Select device to see near real time reports Device CPU Utilization Hum of Swap Capacity Avg Page Faults Processes bas SAD SI a leyte 30 58 160 25 30 42 55199 79 nimitz 30 54 160 54 30 63 55760 63 yorktown 20 96 169 58 30 79 31294 29 laffey 20 92 161 38 31 04 31422 38 hornet 6 88 161 79 13 50 12416 75 eisenhower 5 63 157 79 11 92 12488 17 enterprise 5 58 160 58 12 08 12399 63 Double click on a system in the table to update the content of the following drill down components with performance data for the system The Number of Processes line graph shows the number of processes for the selected system for each sample taken during the report period Number of Processes was
255. s 2 nae ene baw gee nd SALES 6 11 Top len Reports cheer dae eta hid ae op eee ding 6 15 Service Level Management Report 000 eee ae 6 17 Data SOUICE aisha sansa baie heed wa binned eee aE 6 21 Basic Metric Calculations auaa cee eee 6 27 Bay Router ReportPack 0 0c e eee eee eee eee eee eee 7 1 Report Descriptions erob e oe ea a cee eee ee 7 1 Executive Summary Report 00 cece eee eee 7 2 Forecast Report scfcaced tred eenen aE RU od rae SAA bles 7 6 QuickView and Snapshot Reports 000 eee ee eee 7 9 Near Real Time QuickView and Snapshot Reports 7 13 Capacity Planning Report 00 e eee eee ee 7 14 Hot Spots Report Yee erurki tue E aae NEE ees 7 18 Top Ten Report a Eoee E n a En 7 21 Service Level Management Report 00 ee eae 7 23 Data SOUNCE aisaen eit aE E EERE Sail dat te bee BE 7 26 Basic Metric Calculations o neriesi eran cee eee 7 28 Grade of Service Calculation 0 0 c eee eee 7 30 Cisco Router ReportPack 002 e cece eee eee eee 8 1 Report Descriptions 0 0 eae 8 2 Executive Summary Report 0 cece eee eee 8 2 Forecast Report ssc eres erian cece ee teens 8 6 QuickView and Snapshot Reports 2000 eee eee 8 10 QuickView Report 00 eee eee eee 8 10 Snapshot Report 000 e eee eee tee ee 8 14 Near Real Time QuickView and Snapshot Reports
256. s 9 3 QuickView 9 3 RMON2 ReportPack description of 10 1 installing 1 2 report descriptions 10 1 Hot Spots 10 2 QuickView 10 5 Snapshot 10 5 Top Ten 10 9 Router ReportPack data source variables 6 21 dependencies 1 2 description of 6 1 installing 1 2 metrics for 6 27 report descriptions 6 2 Executive Summary 6 2 Forecast 6 5 Hot Spots 6 11 QuickView 6 8 Service Level Management 6 17 Snapshot 6 8 Top Ten 6 15 Service Level Management report description of 2 5 for Cisco Router ReportPack 8 26 for Frame Relay ReportPack 4 20 for LAN ReportPack 5 23 for Router ReportPack 6 17 for System ReportPack 3 41 12 32 for WAN ReportPack 5 23 Snapshot report description of 2 5 for Bay Router ReportPack 7 9 for Cisco Router ReportPack 8 10 for Frame Relay ReportPack 3 20 4 7 for LAN ReportPack 5 11 for RMON2 ReportPack 10 5 for Router ReportPack 6 8 for System ReportPack 12 13 for WAN ReportPack 5 11 snapshot report selecting targets elements for 1 27 System ReportPack 12 1 data source variables 12 33 12 50 Grade of Service calculation for 12 51 installing 1 2 metrics for 12 50 report descriptions 12 1 Capacity Planning Filesystems 12 18 Capacity Planning Memory 12 23 Executive Summary 12 2 Forecast Filesystems 12 4 Forecast Memory 12 8 Hot Spots 12 25 Near Real Time Quick View 12 14 Near Real Time Snapshot 12 14 QuickView 12 9 Service Level Management 3 41 12 32 Snapshot 12
257. s finally sur faced This example shows that the ratio of errors and discards to total packets is fair ly stable over time Daily Errors and Discards lincoln WAN Link Lincoln Tulsa Tue Nov 09 1999 Mon Dec 20 1999 0 NovO9 Nov 16 Novw23 Nov30 DecO Dec14 95th Percentile In Errors 95th Percentile Out Erors 965th Percentile In Discards 95th Percentile Out Discards Figure 3 14 Capacity Planning Daily Errors and Discards Drill Down WAN swioday wo gt lt S gt Ko J m z s Top Ten The Top Ten report identifies the top ten elements with the highest ranked metric values The top ten elements are listed by highest rank and highest change of rank For example the Top Ten report for the Bay Router ReportPack lets the network manager know at a glance which Bay routers produce the greatest throughput and Top Ten 3 17 3 Analyzing TREND Reports exhibit the poorest health The top ten Bay routers ranked by total throughput packets are listed by highest rank as well as highest change of rank Throughput and throughput change charts alert you to the routers greatest changes in throughput suggesting potential capacity overloading if increasing trends continue The throughput change report also may highlight the newly installed devices and technologies that are approaching a logical utilization threshold or are underutilized The following components comprise the Top Ten report for Bay Rou
258. s per second and current day s rank previous day s rank and change in rank from the previous day A large change in rank identifies a volatile resource and you should investigate the reasons for the high change in throughput Largest Bay Routers Throughput Changes Sun Dee 19 1999 Router Pct Discards Pkts sec Rank Change Rank PrevPkts sec Prev Rank Sanjose 3 73 27538522 93 4 1 273397 10 09 6 sanluis 3 48 27366425 47 6 7 27378710 53 5 The Bay Routers with the Most Percentage of Discards summary table ranks Bay routers in descending order by the highest percentage of packet discards for the report day Packets second and bytes second for the selected Bay routers are also shown Bay Routers with Most Percentage of Discards Sun Dec 19 1999 Device Pct Pkts Pkts sec Rank Bytes sec Prior Rank Discarded Sanjose 3 73 27538522 93 3 1182195575 79 3 sanluis 3 48 27366425 47 4 8534497 19 02 4 The Bay Routers with the Greatest Change in Percentage of Discards summary table ranks the selected routers in ascending sequence by change in discard percentage from the previous day A large change in rank identifies a 7 22 Report Descriptions ReportPack Guide volatile resource and you should investigate the reasons for the high change in discard percentage Bay Routers With Greatest Change in Percentage of Discards Sun Dec 19 1999 Device Pct Pkts Pkts sec Rank Bytes sec Change Rank Previous Pct Prior Rank Discarded Pkts Di
259. s the following components The Overutilized Bay Routers summary table shows the Bay routers that are projected to be overutilized within the next 90 days ranked in descending order by the 90 day projected CPU memory or buffer utilization value All calcula tions are based on 95th percentile utilization values and are displayed as per centages in this table Table 7 2 describes how CPU memory and buffer utilization is computed Overutilized means that the router has exceeded the CPU 85 memory 80 or buffer 80 utilization threshold defined Overutilized Bay Routers Projected to Exceed Utilization Threshold Within 90 Days Chresholds CPU Utilization 80 Memory Util 75 Buffer Util 75 Router Slot 30 Day CPU 60 Day CPU 90 Day CPU DTT CPU Util 30DayMemory 60DayMemory 90 Day Util Forecast Util Forecast Util Forecast Forecast Util Forecast Util Forecast Util F sanjose 2 96 58 96 54 96 50 NA 100 00 100 00 AC sanjose 5 92 32 92 27 92 23 N A 100 00 100 00 AC sanjose 3 91 44 91 42 91 39 N A 100 00 100 00 AC sanjose 41 93 86 94 07 94 27 1000 00 96 70 97 58 9 sanjose 4 85 79 86 01 86 23 78 00 92 15 92 28 9g sanluis 3 9 69 9 69 9 69 N A 91 51 91 74 9 sanluis 1 7 80 7 80 7 80 NIA 100 00 100 00 AC U Double click on a Bay router in this table to update the following components with performance data for the selected router The Bay Router Daily Grade of Service stacked bar graph shows GOS scores for CPU memory and buffer
260. sage and Size line graph shows the number of kilobytes for the daily average and maximum transaction log utilization for the last 90 days It also shows the transaction log size in kilobytes l Transaction Log Usage and Size x Average and Maximum Last 90 Days m Wed Apr 28 1999 Mon Jul 26 1999 e E 374 491 60 ay 9 oD 280 868 70 xo n Q 2 O 187 245 80 J 93 622 90 o 0 Apr 28 May12 May26 Jun09S Jun23 JulO 7 Jul 21 Daily Average Usage ULTRA_SYBASE 0 Daily Maximum Usage ULTRA_SY BASE 0 Transaction Log Size ULTRA_SY BASE 0 Report Descriptions 11 5 11 TREND Database ReportPack The Transaction Log Usage bar chart shows the average number of kilobytes of the transaction log used for the baseline period and the number of kilobytes the transaction log will use for 30 60 and 90 day forecasts Transaction Log Max Usage 30 60 and 90 Day Forecast Baseline Tue Jun 15 1999 Mon Jul 26 1999 14 659 80 9 773 20 Kilobytes 4 886 60 o 00 00 00 Wi Baseline ULTRA_SYBASE 0 30 Day ULTRA_SYBASE 0 60 Day ULTRA_SYBASE 0 90 Day ULTRA_SYBASE 0 Even though the Utilization components look very similar to the Usage components they actually show different information The Utilization components change when the database or transaction log sizes change but the Usage components stay the same since they show the actual use Data Source The Usage Summary report uses the daily and fo
261. scarded sanluis 3 48 27366425 47 4 8534497 19 02 4 3 48 4 sanjose 3 73 27638522 93 3 1182195575 79 6 3 73 3 Service Level Management Report The Service Level Management report lets CIOs CFOs network managers and customers know if their Bay routers are meeting contracted service levels for availability and response times by reporting Bay router availability across the enterprise over time and for individual Bay routers for the day It reports the ten Bay routers with lowest availability for the day and the ten routers with the highest network response times It also provides drill down reports for daily availability over the baseline period and daily and hourly response times for the selected routers Routers With Lowest Availability Sun Dee 19 1999 Device Availability sanluis 100 00 with lowest availability or highest network response time NI D The Service Level Management report comprises the following components lt ya The Availability table lists the Bay routers with the lowest availability for the eo report day xR es D Availability 3 Double click on a Bay router in this chart to update the following components with performance data for the selected router Report Descriptions 7 23 7 Bay Router ReportPack The Daily Availability line graph shows the average availability and service level of the selected Bay router for each sample over the report period 42 days Daily Houry Daily Availabi
262. st_id region_id Unique integer value for each integer region region_name Region name associated with varchar 128 region_id value dsi_descr A description of the router varchar 128 Router ReportPack A 5 sewo gt v e me s gt lt at A Property File Formats TREND A 6 Index A Availability definition of 2 7 Baseline definition of 2 10 Bay Router ReportPack data source variables 7 26 dependencies 1 2 description of 7 1 Grade of Service calculation for 7 30 installing 1 2 metrics for 7 28 report descriptions 7 1 Capacity Planning 7 14 Executive Summary 7 2 Forecast 7 6 Hot Spots 7 18 TREND Near Real Time Quick View 7 13 Near Real Time Snapshot 7 13 Quick View Snapshot 7 9 Service Level Management 7 23 Top Ten 3 17 7 21 report descriptionsTop Ten 3 18 C Capacity Planning report description of 2 5 for Bay Router ReportPack 7 14 for Cisco Router ReportPack 8 19 for Frame Relay ReportPack 4 12 for LAN ReportPack 5 16 for WAN ReportPack 3 13 5 16 Capacity Planning Filesystems report for System ReportPack 12 18 Capacity Planning Memory report for System ReportPack 12 23 Index 1 Index Index 2 Cisco Router ReportPack data source variables 8 30 dependencies 1 2 description of 8 1 Grade of Service calculation for 8 36 installing 1 2 metrics for 8 34 report descriptions 8 2 Capacity Planning 8 19 Executive Summary 8 2 Forecast 8
263. t Descriptions 8 29 8 Cisco Router ReportPack The Hourly Response Time line graph shows the response time for each sample for the report day for the selected Cisco router Daily Hourly Hourly Response Time houston Acme s Support Office Sun Dec 19 1999 1 216 00 7 972 00 729 00 486 00 milliseconds 243 00 2S ttt t a 02 00 05 00 08 00 11 00 14 00 17 00 20 00 23 00 Data Source Source data for the Cisco Router ReportPack is collected using SNMP from the Cisco 10 MIB Table 8 1 lists the variables that are polled by SNMP from the Cisco 10 MIB Table 8 1 Cisco Router Variables Variable OID String bufferSmSize 1 3 6 1 4 1 9 2 1 14 Contains the size of small buffers bufferSmTotal 1 3 6 1 4 1 9 2 1 15 1 of 5 8 30 Data Source ReportPack Guide Table 8 1 Cisco Router Variables Variable OID String Contains the total number of small buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferSmFree 1 3 6 1 4 1 9 2 1 16 Contains the number of free small buffers Changed from INTEGER to Gauge by DeskTalk Systems Inc bufferSmHit 1 3 6 1 4 1 9 2 1 18 Contains the number of small buffer hits Changed from INTEGER to Counter by DeskTalk Systems Inc bufferSmMiss 1 3 6 1 4 1 9 2 1 19 Contains the number of small buffer misses Changed from INTEGER to Counter by D
264. t Utilization calculated two ways one using CIR and one using ifSpeed of all samples collected for each hour of the report day Hourly Utilization All P Cs Wed Dec 22 1999 123 20 92 40 61 60 Percent 30 80 0 00 00 06 00 12 00 18 00 00 00 In Utilization CIR Out Utilization CIR In Utilization Ifspeed Out Utilization Ifspeed The Hourly Congestion line graph shows total frames with Backward Explicit Congestion Notification BECN and Forward Explicit Congestion Notification FECN for all PVCs over each hour of the report day Hourly Congestion All P Cs Wied Dec 22 1999 900 00 675 00 460 00 Number of Frames 225 00 0 00 00 06 00 12 00 18 00 00 00 Frames w BECNs Frames wFECNs yoed Report Descriptions 4 5 A TI o 3 D D o o lt D me fe 1 4 Frame Relay ReportPack The Hourly Exception Count stacked bar chart shows at a glance how many PVCs experience exception conditions for utilization and congestion over the day It also includes individual plots of Utilization and Congestion over the day Hourly Exception Count All P Cs Wed Dec 22 1999 24 20 e 15 90 a 3 x W 40 60 o 3 E 530 2 z 0 Wed 00 00 Wed 09 00 Wed 18 00 E Congestion E Out utilization Win Utilization Forecast Report 4 6 The Forecast report lists all PVCs that are within 90 days of reaching a utilization or congestion threshold Those P
265. t initialized A context switch occurs each time a process gives up the CPU and another takes its place This counter reflects the level of system activity a high rate of context switching is indicative of the system load numTraps 1 3 6 1 4 1 546 1 1 7 8 15 The total number of traps switching a process from user to kernel mode that have occurred since the kernel was last initialized Traps are hardware exceptions that occur and are usually caused by the currently running process Example traps include those for incorrect arithmetic operations segmentation violations or page faults An excessive trap rate can indicate a faulty program excessive system loading or some other abnormality P numSyscalls 1 3 6 1 4 1 546 1 1 7 8 16 d ry 4 of 9 p xo Q Data Source al ay pa 12 System ReportPack 12 38 Table 12 1 System Variables Polled from Empire kernelperf Table Variable OID String The total number of system calls that have occurred since the kernel was last initialized System calls occur when a process calls a function that must execute in kernel or privileged mode it is a indicator of the system load and may be higher on servers or machines shared by many users numInterrupts 1 3 6 1 4 1 546 1 1 7 8 17 The total number of device interrupts that have occurred since the kernel was last initialized Devices and peripherals interrupt the operating system wh
266. t period for the selected router Buffer Faults houston Acme s Support Office Mon Dec 20 12 00 AM Mon Dec 20 12 15 PM 3 00 N p 3 5 Number of Faults iv 0 60 0 00 00 02 00 04 00 06 00 08 00 10 00 12 00 The Physical Memory Utilization line graph shows the percentage of memory utilization for each sample taken during the report period for the selected router Physical Memory Utilization houston Acme s Support Office Mon Dec 20 12 00 AM Mon Dec 20 12 15 PM 98 00 78 40 58 80 Percent 39 20 19 60 0 00 00 02 00 04 00 06 00 08 00 10 00 12 00 8 18 Report Descriptions ReportPack Guide Capacity Planning Report The Cisco Router Capacity Planning report shows the most overutilized and underutilized Cisco routers It enables you to allocate the traffic load to achieve maximum Cisco router utilization potential The Capacity Planning report comprises the following components The Overutilized Cisco Routers table ranks the Cisco routers forecasted to reach 95 CPU Utilization 95 Memory Utilization or 80 Buffer Utilization within the next 30 60 or 90 days in descending order Overutilized Cisco Routers Projected To Exceed Utilization Thresholds Within 90 Days Thresholds CPU Utilization 95 Memory Util 75 Buffer Util 75 Router DTTCPU 30DayCPU 60DayCPU 90DayCPU DTT Memory Util 30DayMem 60DayMem 9 Utilization Util Forecast Util Forecast Util Forecast Util Forecast Util For
267. t space on UNIX and C TREND on Windows NT If you have downloaded the ReportPacks from the DeskTalk web site then you have to uncompress the ReportPack archive file For more information about unzipping the compressed archive file see Unzipping the ReportPack Sources on page 1 3 Unzipping the ReportPack Sources Note Before you uncompress the ReportPack sources ensure that you remove or rename any previous versions of the ReportPacks that reside in the DPIPE_HOME packages directory To uncompress the ReportPack archive file 1 If you do not have an unzip utility installed you can download one by entering the following URL in your web browser Installing ReportPacks 1 3 1 ReportPack Administration http www cdrom com pub infozip UnZip html 2 Unzip the ReportPack archive file to the DPIPE_HOME packages directory on the system where you have TREND installed and ensure that the option that preserves the directory and path names is selected 3 Verify that you now have a folder for the ReportPack under the DPIPI E packages folder 4 Use AutoPilot to install the ReportPacks For more information about using AutoPi ReportPack lot see Running Autopilot on page 1 10 Directory Structure The V4 0 Bundled ReportPack suite has a directory structure which did not exist for previous versions A ReportPack may have different installation options For exam ple the Repo
268. tem Filesystem summary table shows the systems that are projected to be the least utilized within the next 90 days It is ranked in ascending sequence by the 90 day projected filesystem utilization value All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Underutilized means that the system is furthest from reaching the filesystem utilization thresholds 30 defined Underutilized Filesystems Device Filesystem Current Filesystem Projected Projected Projected Utilization Util 30 Days Util 60 Days Util 90 Days yortown exporthome1 12 00 12 00 12 00 12 00 Double click on a system in this table to update the following components with performance data for the selection The Grade of Service stacked bar chart shows GOS scores for queue length and CPU memory and swap utilization over the baseline period for the selected system See Table 12 7 for a description of GOS weighting factors The Filesystem Utilization line graph shows the average filesystem utilization over each day of the baseline period for the selected system Report Descriptions ReportPack Guide The CPU Utilization line graph shows the average CPU utilization over each day of the baseline period for the selected system The Memory Utilization line graph shows the average memory utilization over each day of the baseline period for the selected system The Swap Utilization line graph shows the a
269. ter Busiest Bay Routers summary table see Busiest Bay Routers on page 3 18 Largest Bay Routers Throughput Changes summary table see Largest Bay Router Throughput Changes on page 3 19 Bay Routers with Most Percentage of Discards summary table see Bay Routers with the Most Percentage of Discards on page 3 19 Bay Routers with Greatest Change in Percentage of Discards summary table see Bay Routers with the Greatest Change in Percentage of Discards on page 3 20 The following sections describe these charts in more detail For detailed information about the Bay Router ReportPack see Bay Router ReportPack on page 7 1 Busiest Bay Routers 3 18 Figure 3 15 shows the Busiest Bay Routers summary table which lists up to ten Bay routers with the highest throughput packets for the report day Busiest Bay Routers Sun Dec 19 1999 Router Pct Discards Pkts sec Rank Bytes sec Prev Rank sanjose 3 73 27538522 93 4 1182195575 79 6 sanluis 3 48 27366425 47 6 8534497 19 02 5 Figure 3 15 Top Ten Busiest Bay Routers Top Ten ReportPack Guide Largest Bay Router Throughput Changes Figure 3 16 shows the Largest Bay Router Throughput Changes summary table which ranks the selected routers in ascending sequence by change in throughput packets Largest Bay Routers Throughput Changes Sun Dec 19 1999 Router PctDiscards Pkts sec Rank Change Rank PrevPkts sec Prev Rank sanjose 3 73 27538
270. the selected Cisco router for each sample taken on the report day The Percent Discards bar chart shows the percentage of discards for the selected Cisco router for each sample taken on the report day For more information about how the percentage of discards is calculated see Table 7 2 The Buffer Faults bar chart shows the total number of Buffer Faults for the selected Cisco router for each hour of the report day The Exception Detail table shows the number of CPU Utilization Memory Utilization Buffer Utilization and Buffer Faults exception conditions occurring for each sample period on the report day for the selected Cisco router A CPU Utilization gt 95 Memory Utilization gt 75 or Buffer Utilization gt 80 metric greater than 80 is treated as an exception and is included in the chart For more information about how utilization and buffer faults are calculated see Table 8 2 Report Descriptions ReportPack Guide Top Ten Report The Top Ten report lets the network manager know at a glance which Cisco routers have the highest throughput and discard percentage The top ten routers ranked by throughput and discard percentage are listed by highest rank as well as highest change of rank The Cisco Routers with Highest Throughput table shows the ten Cisco routers with the highest throughput measurements for the report day The Cisco router with the highest throughput is listed first The Cisco router s throughput
271. the Grade of Service metric see Table 2 2 on page 2 6 For more information about Grade of Service as it applies to each type of report see Analyzing TREND Reports on page 3 1 4 28 Grade of Service Calculation TREND 5 LAN and WAN ReportPacks The LAN Local Area Network and WAN Wide Area Network Connectivity ReportPacks direct polling of data from routers and other devices that support MIB II to generate comprehensive report sets for the following Half duplex interfaces such as Ethernet FDDI and Token Ring interfaces LAN Full duplex interfaces such as T 3 and DSL WAN The LAN and WAN Connectivity ReportPacks answer questions including Do the LAN or WAN networks provide an acceptable level of service to customers or end users Are there overloaded LAN or WAN interfaces If so are there underutilized LAN or WAN interfaces that can help balance the load Are performance problems due to high levels of congestion errors excessive utilization or other causes What are the top ten LAN or WAN interfaces contributing to the poor health of the network Which LAN or WAN interfaces are likely to degrade service in the near future if preventive action is not taken 5 1 syoequoday oO g gt z o a gt z 5 LAN and WAN ReportPacks Report Descriptions The following sections describe each report in the LAN and WAN ReportPacks in detail Executive Summary Report 5 2
272. the Quick View report however it allows you to show detailed information about a particular Cisco router The selected Cisco router may be different than the Cisco routers appearing in the QuickView report QuickView Report The QuickView report comprises the following components Report Descriptions ReportPack Guide The Grade of Service stacked bar chart shows the Cisco routers with the highest daily GOS scores and 95th percentile of the GOS score for Memory Utilization CPU Utilization Buffer Utilization and Buffer Faults Table 8 3 describes how Grade of Service is calculated Grade Of Service Cisco Routers With Highest Grade Of Service GOS Scores Sun Dec 19 1999 Critical Warning Good Excellent Di butter Faults E Buffer utilization CPU utilization Wiemory Utilization BOINO JPUUOSJ J SPUN uojsoq BOINO Sales saw epo aoyo poddng saway uonoy poddns Jiepisag ujoouly pyoddns 4 epseq esin Double click on a router in the chart to update the content of the following components with performance data for the selected router The Hourly Volume bar chart shows the volume of the selected Cisco router for each hour of the report day For more information about how Volume is computed see Table 7 2 Hourly Yolume boston Acme s Personnel Office Sun Dec 19 1999 4 975 60 3 731 70 2 487 80 Megabytes 1 243 90 0 Sun 00 00 Sun 09 00 Sun 18 00 Report Descriptions 8 11 D o S e
273. thly average and peak utilization and average and peak percentage collisions The interfaces are sorted in descending order based on monthly peak utilization Percent collisions is calculated by dividing the number of collisions by the number of good packets Report Descriptions 9 5 9 RMON ReportPack Monthly RMON Ethernet Statistics Sorted by Peak Utilization Pug 1999 RMOH Probe LF Avg Pct Util Peak Pct Util Avg Pct Peak Pct Collisions Collisions sanford 1_ALL 3 22 27 19 0 02 0 07 fms east desktalk com 1_ALL 178 22 67 0 01 0 02 fp2000 east desktalk com 1_ALL 1 73 21 57 2 15 0 74 fp2000 east desktalk com 2_ALL 1 70 21 48 0 47 0 37 bluegrass 3_ALL 0 32 1 08 0 00 0 00 bluegrass 110_Bl_ALL 0 32 1 08 0 00 0 00 bluegrass 110_DTE_ALL 0 17 0 78 0 00 0 00 bluegrass 21_ALL 0 17 0 78 0 00 0 00 Double click on an interface in the table to update the content of the following components with performance data for the selected interface The RMON Ethernet Utilization bar chart shows the monthly average 95th percentile and peak utilization for the past 12 months if available for the interface selected in the Monthly RMON Ethernet Statistics table RMON Ethernet Utilization bluegrass 1_ALL Pug 1999 Aug 1999 56 50 46 20 e 2 N 33 90 z T 22 60 3 a a 11 30 o 20 00 00 E Avg Pot Util EM Peak Pct Util 9 6 Report Descriptions ReportPack Guide The RMON Ethernet Collisions bar chart shows the monthly average and p
274. tilization Figure 3 32 Hot Spots CPU Utilization Drill Down Cisco Router By showing the time of day when an exception occurs the drill down components can help narrow possible causes of the problem For example a consistent spike in utilization at 11 00 a m might indicate that a resource intensive application such as a backup is being run In that case you can schedule the application to run at some other time such as 11 00 p m to balance the router load Figure 3 32 shows CPU utilization is following the normal workday hours 3 36 Hot Spots Report ReportPack Guide Buffer Utilization Drill Down Figure 3 33 shows the Buffer Utilization line graph which displays the percentage of the various size buffers Huge Buffer Big Buffer Large Buffer Medium Buffer and Small Buffer that are in use during each time interval over the past 24 hours Buffer Utilization houston Acme s Support Office Sun Dee 19 1999 400 00 ae e 0 ee o S o 00 00 06 00 12 00 18 00 00 00 Percent Huge Buffer Utilization Big Buffer Utilizaiton Large Buffer Utilization Medium Buffer Utilization Small Buffer Utilization Figure 3 33 Hot Spots Buffer Utilization Drill Down Cisco Router swioday wo gt Ss lt S gt Q J m z s Hot Spots Report 3 37 3 38 3 Analyzing TREND Reports Memory Utilization Drill Down Figure 3 34 shows the Memory Utilization line graph Memory u
275. tilization CPU Utilization Buffer Utiliza tion and Buffer Errors for each hour of the report day Table 8 3 describes how Grade of Service is calculated The Hourly Volume bar chart shows the volume of the selected Cisco router for each hour of the report day For more information about how Volume is computed see Table 7 2 The CPU Utilization Free Memory and Total Buffers pie charts show the percentage of samples taken during the day in which the CPU Utilization Free Memory Utilization and Total Buffers metrics calculated for each sample fall in the defined range The Percent Errors line graph shows the total percentage of errors for each hour during the day For more information about how percent errors is computed see Table 7 2 The Percent Discards line graph shows the percentage of discards for each hour during the day For more information about how Discards is computed see Table 7 2 Near Real Time QuickView and Snapshot Reports The Near Real Time Quick View report provides router performance statistics up to the last SNMP poll The report does not rely on nightly summaries therefore it provides instant reporting on collected data Network managers and analysts can go to this report to select a router from a table and assess its status in the following performance areas CPU utilization buffer utilization and memory utilization Alternatively the Snapshot report shows the same information however you sel
276. tilization is consistently high indicating that a memory upgrade is required Memory Utilization houston Acme s Support Office Sun Dee 19 1999 98 70 65 80 Percent 32 90 00 00 06 00 12 00 18 00 00 00 95th Percentile Memory Util Average Memory Utilization Figure 3 34 Hot Spots Memory Utilization Drill Down Cisco Router Hot Spots Report ReportPack Guide Percent Discards Drill Down Figure 3 35 shows the Percent Discards bar chart for each hour of the day If a high discard percentage correlates with a high CPU utilization number the router is having trouble keeping up with the traffic volume Percent Discards houston Sun Dee 19 1999 3 00 Percent 1 00 S n 00 00 Sun 12 00 Mon 00 00 Figure 3 35 Hot Spots Percent Discards Drill Down Cisco Router Buffer Faults Drill Down Figure 3 36 shows the Buffer Faults bar chart which displays the total number of buffer faults for each hour of the day If a high number of buffer faults correlates Hot Spots Report 3 39 suyod y wo gt Ss lt S gt Ko J m z oO 3 Analyzing TREND Reports with times of high CPU utilization the router is having trouble keeping up with the traffic volume Buffer Faults houston Acme s Support Office Sun Dee 19 1999 12 00 8 00 Total 4 00 S n 00 00 Sun 09 00 Sun 18 00 Figure 3 36 Hot Spots Buffer Faults Drill Down Cisco Router Exception Detail Drill Down
277. tion physical_mem_util 1 freememory 1 physicalram 100 The ratio of total memory in bytes used to total memory available in bytes is multiplied by 100 to yield a percentage Buffer Utilization memp1_util memp1_free memp1_count 100 0 small memp2_util memp2_free memp2_count 100 0 medium memp3_util memp3_free memp3_count 100 0 large memp4_util memp4_free memp4_count 100 0 big memp5_util memp5_free memp5_count 100 0 huge The ratio of the number of free buffers to the total number of buffers times 100 to yield a percentage D o me fe U o P x JOINOY ODSID 8 Basic Metric Calculations 8 35 8 Cisco Router ReportPack Grade of Service Calculation Table 8 3 lists how the metrics are weighted in Cisco Router Grade of Service GOS charts Table 8 3 Cisco Router GOS Calculations GOS CPU Memory Buffer Calculation Utilization Utilization Utilization Buffer Faults Weighting 10 40 20 30 Excellent 0 19 0 19 0 19 0 99 Score Range Good 20 39 20 39 20 39 100 399 Score Range Warning 40 79 40 79 40 79 400 799 Score Range Critical 80 100 80 100 80 100 800 and over Score Range For more information about the Grade of Service metric see Table 2 2 on page 2 6 For more information about Grade of Service as it applies to each type of report see Ana
278. tistics based on the data polled for the last 6 hours The statistic shown for each metric is the average of all the data collected for that statistic over the 6 hour period yoed Double click on a device in the table to update the content of the following drill down components with performance data for the device N ius o lt D e i Po xe e Di The CPU Utilization line graph shows the CPU utilization for each sample taken during the report period for the selected router The Buffer Utilization line graph shows the buffer utilization for each sample taken during the report period for the selected router The Memory Utilization line graph shows the memory utilization for each sample taken during the report period for the selected router Report Descriptions 7 13 7 Bay Router ReportPack Capacity Planning Report 7 14 The Capacity Planning report shows the most overutilized and underutilized Bay routers that are within 90 days of reaching a CPU 85 memory 80 or buffer utilization 80 threshold It allows the CIO and network manager to estimate utilization levels in 30 60 or 90 days You can double click on any interface in the Overutilized Bay routers or Underutilized Bay routers summary tables to display corresponding drill down charts that detail daily data for Grade of Service and CPU memory and buffer utilization for the baseline period The Capacity Planning report comprise
279. ttention on the growth and changes over the last 24 hours Volume Baseine Hourly Volume All LAN Interfaces Mon Dec 20 1999 3 30 Hed iy Gigabytes o 0 Mon 00 00 Mon 09 00 Mon 18 00 Figure 3 2 Executive Summary Hourly Volume Chart LAN 3 4 Executive Summary Report ReportPack Guide Click on the Baseline tab to see the Hourly Baseline bar chart Figure 3 3 shows the Hourly Baseline chart Volume Baseline Hourly Baseline All LAN Interfaces Tue Nov 09 1999 3 30 N h Gigabytes 0 00 00 06 00 12 00 18 00 Figure 3 3 Executive Summary Hourly Baseline Chart LAN This chart shows the average volume by hour of day over the baseline period The hourly baseline is a much more valuable indicator than a simple average or a daily average baseline Since traffic typically varies by hour of day for example the traffic at 10 00 a m is typically much higher than at midnight TREND calculates individual baseline values for each hour of the day over the baseline period to deliver the most accurate and useful comparisons Comparing this chart to the Hourly Volume chart will allow the executive to see if yesterday s traffic deviated significantly from the historical average swioday wo gt Ss lt S gt Ke J m z oO Hourly Grade of Service Chart Figure 3 4 shows the Hourly Grade of Service stacked bar chart which correlates a
280. ty and daily and hourly response times for these elements Note For an in depth analysis of how to interpret the System Service Level Management report and for chart illustrations see Service Level Management Report on page 3 41 The Service Level Management report comprises the following components The Availability list shows the systems with the lowest availability for the report day Report Descriptions ReportPack Guide Double click on a system in this list to update the following component with performance data for the selected system The Daily Availability line graph shows the systems with the lowest availability for the report day compared to the service level for the baseline period The Hourly Availability line graph shows the systems with the lowest availability for each hour of the report day The Response Time list shows the systems with the highest response time for the report day Double click on a system in this list to update the following component with performance data for the selected system The Daily Response Time line graph shows the daily response time for the selected System for each sample over the report period 42 days The Hourly Response Time line graph shows the response time for the selected system for each sample for the preceding day Data Source Source data for the System ReportPack is collected from the Empire Technology s SystemEDGE product The data com
281. udes the CPU s clock speed cpuStatsIdle 1 3 6 1 4 1 546 13 1 1 3 The total number of ticks spent by this CPU in Idle mode This variable is not supported by NT This value may wrap a 32 bit counter cpuStatsUser 1 3 6 1 4 1 546 13 1 1 4 The total number of ticks spent by this CPU in User mode This value may wrap a 32 bit counter 1 of 3 Data Source 12 43 _ N w lt P mp 3 D D O U Q 12 System ReportPack Table 12 3 System Variables Polled from Empire cpuStatsEntry Table Variable OID String cpuStatsSys 1 3 6 1 4 1 546 13 1 1 5 The total number of ticks spent by this CPU in Kernel or system mode This value may wrap a 32 bit counter cpuStats Wait 1 3 6 1 4 1 546 13 1 1 6 The total number of ticks spent by this CPU in Wait mode Wait mode is that time spent waiting on I O or swap It may wrap a 32 bit counter cpuStatsLastUpdate 1 3 6 1 4 1 546 13 1 1 7 The time based on sysUpTime that the counters for this CPU were last updated Because these values are cached periodically by the agent the time of last update is important for determining the interval over which CPU mode percentages are calculated cpuStatsIdlePercent 1 3 6 1 4 1 546 13 1 1 8 The percentage of time over the sample period the system s CPU was idle cpuStatsUserPercent 1 3 6 1 4 1 546 13 1 1 9 The per
282. uide Table 11 2 TREND Database Baseline and Forecast Variables Variable Description AV GAVGpct_dbused The average of daily averages for the percentage of the database used over the baseline period AVGMAXkb_logused The average of the daily maximum number of kilobytes used by the transaction log over the baseline period AVGMA Xpct_logused The average of the daily maximum percentage of the transaction log used over the baseline period DTTAVGpct_dbused The projected number of days to reach the threshold of 80 of the daily average database usage The ReportPack calculates but does not display this variable at this time DTTMAXpct_logused The projected number of days to reach the threshold of 80 of the daily maximum transaction log usage The ReportPack calculates but does not display this variable at this time F30AVGkb_dbused The 30 day projection for the average number of kilobytes the database will use F30AVGpct_dbused The 30 day projection for the average percentage of database use F30MAXkb_logused The 30 day projection for the maximum number of kilobytes the transaction log will use F30MA Xpct_logused The 30 day projection for the maximum percentage of transaction log use F60AVGkb_dbused The 60 day projection for the average number of kilobytes the database will use 2 of 3 Data Source 11 9 Ee m 72 T a
283. utilization over the baseline period Report Descriptions ReportPack Guide for the selected router See Table 7 3 for a description of the GOS weighting factors Bay Router Daily GOS sanjose Desktalk Support 1 Tue Dec 21 1999 Mon Jan 31 2000 Critical Warning Good Excellent Dec21 Jan 04 Jan 18 HE CPU Utilization E buffer utilization Bittemory Utilization The Bay Router Daily CPU Utilization line graph shows the average and 95th percentile CPU utilization over each day of the baseline period for the selected router Bay Router Daily CPU Utilization sanjose Desktalk Support 1 Tue Dec 21 1999 Mon Jan 31 2000 98 00 73 50 Percent o 24 50 0 Dec21 Jan 04 Jan 18 Awg CPU Utilization P95 CPU Utilization Report Descriptions 7 15 yoed N ius o lt D fo 7 Po xe fo Di 7 Bay Router ReportPack The Bay Router Daily Buffer Utilization line graph shows the average and 95th percentile buffer utilization over each day of the baseline period for the selected router Bay Router Daily Buffer Utilization sanjose Desktalk Support 1 Tue Dec 21 1999 Mon Jan 31 2000 98 80 74 10 Percent amp 24 70 0 Dec21 Jan 04 Jan 18 Awg Buffer Utilization P95 Buffer Utilization The Bay Router Daily Memory Utilization line graph shows the average and 95th percentile memory utilization over each day of the baseline period for
284. verage swap utilization over each day of the baseline period for the selected system The Page Faults bar chart shows the number of page faults for each day of the baseline period for the selected system Capacity Planning Memory Report The Capacity Planning Memory report shows the most overutilized and underuti lized systems that are within 90 days of reaching a memory utilization threshold It allows the CIO and network manager to estimate utilization levels in 90 days You can double click on any interface in the Overutilized Systems or Underutilized Sys tems summary tables to display corresponding drill down charts that detail daily data for Grade of Service page faults and CPU memory and swap utilization for the baseline period It shows summaries of the systems that are forecasted to exceed 100 of the memory used on the system in the next 90 days for overutilized and less then 30 of the memory used on the system in the next 90 days for underutilized Note The charts for the Capacity Planning Memory report are identical to the charts for the Capacity Planning Filesystems report For chart illustrations see Capacity Planning Filesystems Report on page 12 18 The Capacity Planning Memory report comprises the following components The Overutilized System Memory summary table shows the computer systems that are projected to be overutilized within the next 90 days ranked in descending order by the 90 day projected
285. vice time in milliseconds for operations served on this disk over the last measurement period This could be expressed as disk busy time number of transfers diskStatsUtilization 1 3 6 1 4 1 546 12 1 1 4 The utilization rate percentage utilization for this disk over the last measurement period This could also be expressed as disk busy time elapsed time 100 diskStatsK BytesTransferred 1 3 6 1 4 1 546 12 1 1 5 The number of KBytes transferred to from this disk diskStatstransfers 1 3 6 1 4 1 546 12 1 1 6 The number of transfer operations to from this disk disksStatsReads 1 3 6 1 4 1 546 12 1 1 7 The number of read operations from this disk diskStats Writes 1 3 6 1 4 1 546 12 1 1 8 The number of write operations to this disk diskStatsHostmibDevTableIndexx 1 3 6 1 4 1 546 12 1 1 9 2 of 3 Data Source ReportPack Guide Table 12 4 System Variables Polled from Empire diskStatsEntry Table Variable OID String The index of this disk device in the device table of the Host Resources MIB diskStatsLastUpdate 1 3 6 1 4 1 546 12 1 1 10 The time in SNMP Time Ticks of the last update of this diskStatsTable entry 3 of 3 Table 12 5 lists the variables from the devTableEntry table in the empire mib file Table 12 5 System Variables Polled from Empire devTableEntry Table Variable OID String
286. with lowest availability and the ten routers with the highest network response times It also provides drill down reports for daily 8 26 Report Descriptions ReportPack Guide availability over the baseline period for the routers with lowest availability and daily and hourly response times for the routers with the highest network response time The Service Level Management Report comprises the following components The Availability table lists Cisco Routers with Lowest Availability for the Day the Cisco routers with the lowest availability for yesterday For more information about how availability is calculated see Daily Availability Drill Down on page 3 42 Availability Routers With Lowest Availability Sun Dee 19 1999 Device Availability tulsa 100 00 cotati 100 00 lincoln 100 00 boston 100 00 houston 100 00 Double click on a Cisco router to update the content of the following components with performance data for the selected router Report Descriptions 8 27 D o S e U o i x J9 NOY ODSID 8 8 Cisco Router ReportPack The Daily Availability line graph shows the average availability and service level of the selected Cisco router for each sample over the report period 42 days By default the service level is set at 99 5 Daily Hour Daily Availability miami Mon Now 08 1999 Sun Dec 19 1999 100 00 80 00 60 00 Percent 40 00 20 00 Nov15
287. wn 12 18 Report Descriptions ReportPack Guide All calculations are based on 95th percentile utilization values and are displayed as percentages in this table Overutilized Filesystems Projected To Exceed 90 Utilization Within 90 Days Device Filesystem Current Filesystem Projected Projected Projected Utilization Util 30 Days Util 60 Days Util 90 Days yortown exporthome 95 00 95 00 95 00 95 00 Double click on a system in this table to update the following components with performance data for the selection The Grade of Service stacked bar chart shows GOS scores for CPU memory and swap utilization and queue length over the baseline period for the selected system See Table 12 7 for a description of the GOS weighting factors Grade of Service GOS Sep 10 Sep 24 E Queue Length OD swap Utilization Dtemory Utilization E CPU Utilization Report Descriptions 12 19 n N w lt P 3 D O U D Q 12 System ReportPack The Filesystem Utilization line graph shows the average filesystem utilization over each day of the baseline period for the selected system Filesystem Utilization 92 10 61 40 a 30 70 o Sep 10 Sep 24 Oct O08 The CPU Utilization line graph shows the average CPU utilization over each day of the baseline period for the selected system CPU Util Memory Util Swap Util Page Faults CPU Utilization wasp Fri Sep 10 1999 Thu Oct 21 1999 6
288. xcellent NovO9 Nov13 Novi Nov21 Now25 Now29 DecO3 DecO Dec11 Dec15 Dec19 E Errors M Discards W Utilization 5 4 Report Descriptions ReportPack Guide The Daily Utilization line graph shows the average and 95th percentile utilization LAN or input and output utilization WAN for the selected interface for each day of the baseline period syoequoday o m gt Z 5 2 gt a Daily Utilization sanjose Sales Tue Nov 09 1999 Mon Dec 20 1999 36 40 23 60 Percent 11 80 Nov 09 Nov 23 DecO7 Awg Utilization 95th Percentile Util The Daily Errors and Discards line graph shows the average percent and 95th percentile errors and discards for the selected interface for each day of the baseline period Daily Errors and Discards sanjose Sales Tue Now 09 1999 Mon Dec 20 1999 6 20 Percent w a 0 Nov 09 Nov 23 DecO Avg Percent Errors 95th Percentile Errors Awg Percent Discards 965th Percentile Discards Report Descriptions 5 5 5 LAN and WAN ReportPacks QuickView and Snapshot Reports 5 6 The QuickView report provides the network management staff with detailed information on key metrics for individual interfaces with the highest GOS scores The Snapshot report provides the same information as the QuickView report for individual interfaces that you select from a pick list for inclusion in the report For more information about selecting a int
289. xcellent Good Warning Critical to each time interval TREND provides an aggregate view of the stability or changes in overall health for the element In Figure 3 31 router health starts at Good and degrades to Warning at 5 00 a m largely due to increased CPU and physical memory utilization The CPU Utilization drill down chart Figure 3 32 provides details on the exact changes that took place You can further investigate buffer utilization in the Buffer Utilization drill down chart Figure 3 33 The physical memory utilization Figure 3 34 is consistently high This indicates a memory upgrade is required The Buffer Faults drill down chart Figure 3 36 reveals a minimal number of faults the faults that do exist correlate to the CPU Utilization increases Together these charts indicate a router that is pushing the limits of its CPU and physical memory resources It is not yet discarding much traffic so end user service levels are not severely impacted but the need for an upgrade is indicated swioday wo gt lt S gt Ke J m z oO Hot Spots Report 3 35 3 Analyzing TREND Reports CPU Utilization Drill Down Figure 3 32 shows the CPU Utilization line graph shows hour by hour changes and trends in CPU utilization for the selected Cisco router CPU Utilization houston Acme s Support Office Sun Dee 19 12 00 AM Mon Dec 20 12 00 AM 0 00 00 06 00 12 00 18 00 00 00 P95 CPU Utilization Awg CPU U
290. y Hourly Availibility miami Sales Mon Dec 20 12 00 Aw 100 00 50 00 Percent 0 00 00 02 00 04 00 06 00 08 00 10 00 12 00 Awailability The LAN Interfaces with Highest Response Time for the Day bar graph shows the interfaces with the highest average response times for the day in descending sequence from left to right Double click on a interface in this chart to update the following charts with detailed information for the selected interface Response Time Highest SNMP Response Time by Interface Mon Dec 20 1999 Device Interface Response Time miami 2 98 00 sanjose 1 98 00 sanjose 2 98 00 cotati 2 98 00 houston 2 98 00 houston 3 98 00 boston 2 98 00 tulsa 2 98 00 sanluis 2 93 00 Report Descriptions 5 25 5 LAN and WAN ReportPacks The Daily Response Time line graph shows the average daily network response time for the selected interface for each day of the baseline period Response time is taken from the database column received_usec which is the time in milliseconds between the first SNMP GET and last GET RESPONSE request If multiple packets are needed to get the entire MIB the time for all packets is included Daily Hourly Response Time miami Sales Tue Nov 09 1999 hion Dec 20 1999 122 80 92 10 61 40 milliseconds 30 70 Nov 09 Nov 23 DecO The Hourly Response Time line graph shows the response time received_usec for each sample for the repo
291. y for the Day on page 3 33 and the following drill down components 3 32 Hot Spots Report ReportPack Guide Hourly Grade of Service chart see Hourly Grade of Service Drill Down on page 3 34 CPU Utilization chart see CPU Utilization Drill Down on page 3 36 Buffer Utilization chart see Buffer Utilization Drill Down on page 3 37 Memory Utilization chart see Memory Utilization Drill Down on page 3 38 Percent Discards chart see Percent Discards Drill Down on page 3 39 Buffer Faults chart see Buffer Faults Drill Down on page 3 39 Exception Detail summary table see Exception Detail Drill Down on page 3 40 The following sections describe these charts in more detail For more information about the Cisco Router ReportPack see Cisco Router ReportPack on page 8 1 Problem Summary for the Day Figure 3 30 shows the Problem Summary for a Day table which summarizes multiple exception conditions into one aggregate total number Total Exceptions indicating the top ten offenders ranked by total number of exceptions over the day swioday Problem Summary for the Day Cisco Routers With Highest Exception Counts Sun Dec 19 1999 Router CPU Utilization Memory Utilization Buffer Utilization Buffer Failure Total Exceptions Exceptions Exceptions Exceptions Exceptions cotati 0 00 46 00 96 00 0 00 141 00 wo gt lt S gt Ko J m z Oo
292. y supplying weighted values for queue length and swap Report Descriptions 12 3 _ N 2 lt P 3 D O U D Q 12 System ReportPack memory and CPU utilization for each hour of the report day Table 12 7 describes the GOS weighting factors used in this chart Hourly Grade of Service All Systems Thu Oct 21 1999 Critical Warning E Queue Length m Swap Utilization Memory Utilization E CPU Ltilization Good Excellent 00 00 02 00 04 00 06 00 08 00 10 00 12 00 14 00 16 00 18 00 20 00 22 00 Forecast Filesystems Report 12 4 The Forecast Filesystems report lists a summary of all filesystems that are within 90 days of reaching a disk space utilization threshold Those filesystems closest to the 90 day threshold appear in the list first The calculation for the threshold is from the baseline period which is the previous 42 days The generation of the forecast data uses the 95th percentile values from the daily exception table The report arms the network manager with the information required to assess performance and take preventive action Each element in the summary hot links to multiple graphs that show CPU memory and swap utilization grade of service and page faults for trend analysis over the baseline period Note that there are usually multiple filesystems for each computer system each filesystem references allocated disk space on the computer The Forecast Filesystems
293. ystem for each day of the baseline period CPU Lil Memory Util Swap Util Page Faults Page Faults wasp C Fri Sep 10 1999 Thu Oct 21 1999 2 570 036 80 1 927 527 60 1 286 018 40 642 509 20 Number of Page Faults oO Sep09 Sep13 Sep17 Sep21 Sep25 Sep29 OctO3 OctO7 Oct11 Oct16 Oct19 Forecast Memory Report 12 8 The Forecast Memory report lists all systems within 90 days of a memory utiliza tion threshold Those systems closest to a threshold appear in the list first The report arms the network manager with the information required to assess performance and take preventive action Each element in the summary hot links to multiple graphs that show CPU memory and swap utilization grade of service total swaps and page faults for trend analysis of the baseline performance period Note The charts for the Forecast Memory report are identical to the charts for the Forecast Filesystems report For chart illustrations see Forecast Filesystems Report on page 12 4 The Forecast Memory report comprises the following components The Estimated Days to Threshold DTT summary table lists the systems that are within 90 days of a 100 percent memory utilization threshold It shows the number of days to threshold for memory utilization current memory Report Descriptions ReportPack Guide utilization and projected memory utilization in 90 days for each computer system Double click on a c

Grade of Service - HP Software Support

Contents

Download Pdf Manuals

Related Search

Related Contents