Guardian Performance Analyzer (GPA) Manual Abstract GPA consolidates and analyzes system performance data collected by Measure, the Tandem performance measuring product. On the basis of its analysis, GPA indicates how well or how poorly the system is performing and makes recommendations for improving the overall performance of the system. Product Version GPA D00 and G00 Supported Releases This manual supports C30.09, D00, and G00 and all subsequent releases until otherwise indicated in a new edition.
Document History Part Number Product Version Published 135081 GPA D00 and G00 January 1998 089503 C30.09 March 1993 New editions incorporate any updates issued since the previous edition. A plus sign (+) after a release ID indicates that this manual describes function added to the base release, either by an interim product modification (IPM) or by a new product version on a .99 site update tape (SUT). Ordering Information For manual ordering information: domestic U.S.
Guardian Performance Analyzer (GPA) Manual Glossary Index Examples What’s New in This Manual vii Manual Information vii Additions and Changes vii About This Manual ix ix Related Manuals Manual Organization ix Your Comments Invited ix Notation Conventions x 1. Introducing GPA GPA Benefits and Features 1-1 How GPA Works 1-2 Standard Report Sections 1-4 Optional Report Sections 1-4 Error Reports 1-5 GPA Requirements 1-5 2.
3. Description of GPA Reports Contents 3.
Glossary Contents Interpreting the Optional Report 4-24 Optional Report Section for the SYSTEM Class 4-25 Optional Report Section for the SERVER Class 4-27 Implementing GPA Tuning Recommendations 4-29 The Detail System Performance Score Reports 4-30 4-31 Negative Contributing Factors to CPU Score Negative Contributing Factors to Memory Subsystem Score 4-32 Disk Cache Performance Score Grading 4-33 Negative Contributing Factors to Disk Volume Score System Recovery Performance Score Grading 4-33
Examples Contents Example 3-14. Process Move Recommendations 3-17 Example 3-15. Expected System Performance After Tuning 3-18 Example 3-16. Cover Page to the Optional Report Example 3-17. Summary Section for the System Class 3-21 Example 3-18. Additional Information in the Detail Section for the System Class 3-22 Example 4-1. Node Characteristics for \NODEA 4-2 Example 4-2. Node Characteristics for \NODEB 4-2 Example 4-3. Node Characteristics for \NODEC 4-3 Example 4-4.
Figures Contents Example 4-24. Process Move Recommendations for \NODEA 4-21 Example 4-25. Process Move Recommendations for \NODEB 4-21 Example 4-26. Primary Switch Recommendations for \NODEB Example 4-27. Expected System Performance After Tuning Changes for \NODEA 4-23 Example 4-28. Expected System Performance After Tuning Changes for \NODEB 4-24 Example 4-29. Cover Page to the Optional Report for \NODEC Example 4-30. Summary Section for the System Class for \NODEC Example 4-31.
Figures Contents Guardian Performance Analyzer (GPA) Manual— 135081 vi
What’s New in This Manual Manual Information Guardian Performance Analyzer (GPA) Manual Abstract GPA consolidates and analyzes system performance data collected by Measure, the Tandem performance measuring product. On the basis of its analysis, GPA indicates how well or how poorly the system is performing and makes recommendations for improving the overall performance of the system. Product Version GPA D00 and G00 Supported Releases This manual supports C30.
What’s New in This Manual Guardian Performance Analyzer (GPA) Manual— 135081 viii
About This Manual This manual describes the Guardian Performance Analyzer (GPA) and tells you how to install and use it on a Tandem system. The manual is intended mainly for system performance analysts, system managers, and others responsible for the performance of Tandem systems. We assume in this manual that you are familiar with the Guardian operating system and with Measure, the Tandem system performance measuring product.
Notation Conventions About This Manual Notation Conventions General Syntax Notation The following list summarizes the notation conventions for syntax presentation in this manual. UPPERCASE LETTERS. Uppercase letters indicate keywords and reserved words; enter these items exactly as shown. Items not enclosed in brackets are required. For example: MAXATTACH lowercase italic letters. Lowercase italic letters indicate variable items that you supply. Items not enclosed in brackets are required.
Notation for Messages About This Manual An ellipsis immediately following a single syntax item indicates that you can repeat that syntax item any number of times. For example: "s-char..." Punctuation. Parentheses, commas, semicolons, and other symbols not previously described must be entered as shown. For example: error := NEXTFILENAME ( file-name ) ; LISTOPENS SU $process-name.
Notation for Management Programming Interfaces About This Manual [ ] Brackets. Brackets enclose items that are sometimes, but not always, displayed. For example: Event number = number [ Subject = first-subject-value ] A group of items enclosed in brackets is a list of all possible items that can be displayed, of which one or none might actually be displayed.
Change Bar Notation About This Manual The message types specified in the REPORT clause are different in the COBOL85 environment and the Common Run-Time Environment (CRE). The CRE has many new message types and some new message type codes for old message types. In the CRE, the message type SYSTEM includes all messages except LOGICAL-CLOSE and LOGICAL-OPEN.
Change Bar Notation About This Manual Guardian Performance Analyzer (GPA) Manual— 135081 xiv
1 Introducing GPA The Guardian Performance Analyzer (GPA) is a software tool designed specifically for use by system performance analysts, system operations managers, or other persons responsible for the proper performance of a Tandem system. GPA consolidates and analyzes system performance data collected by Measure, the Tandem performance measuring product.
How GPA Works Introducing GPA • • • • Provides PUP or SCF input (PUPIN/SCFIN) control statements for implementing tuning recommendations and a PUP or SCF backout file (PUPBAK/SCFBAK) to restore the system to its previous state, depending on the operating system. Enables you to choose the optional reports that meet your specific requirements. Analyzes any Tandem node with up to 16 processors of any type.
How GPA Works Introducing GPA Figure 1-1. How GPA Works Measure Data MEASFH CPU PROCESS DISC OBEYTUNE TUNEDATA GPA #OUT Optional Standard Errors PUPIN / SCFIN PUPBAK / SCFBAK CDT 001.CDD The GPA module is essentially a model of a Tandem node. In performing its analysis, GPA breaks the system down into the following subsystems: CPU, memory, disk volume, and disk cache.
Standard Report Sections Introducing GPA PUPBAK is a PUP INPUT file that can be run after PUPIN, if necessary, to restore the system to its original measured state. SCFIN and SCFBAK have the same function as PUPIN and PUPBAK, but the PUP utility is used on systems prior to G02. SCF is used on G02 and later operating systems. Either PUP files or SCF files will be created, depending on the operating system.
Optional Report Sections Introducing GPA Optional Report Sections In the GPA Optional Report, you specify the level of additional detailed information to be provided. You choose the type of output you want for each report section. Three output types are available: • • Noprint: Skip this optional report section. Summary: A breakdown of the number of processes in each CPU for each of the following reports (each of which is a class of process).
GPA Requirements Introducing GPA GPA requires the use of the following Tandem products: Item Tandem Product Number Measure* T9086 SORT** or FASTSORT** (part of Guardian) * Must be installed on the system being analyzed. ** Must be installed on the system on which GPA is run.
2 Running GPA This section begins with an overview of the procedure for running GPA, followed by detailed information about the procedure. We strongly recommend that you read through the overview section before you attempt to install or run GPA for the first time. Note that using GPA also involves the use of the Tandem Measure performancemonitoring product. The GPA procedure in this section includes the required steps for using Measure in conjunction with GPA.
Running Measure Running GPA an eight-hour window and one-hour interval provides Measure data in one-hour segments. In establishing these periods, keep in mind the following: • • • • • The measurement window depends on how the system is used and should normally equal the daily period of continuous use. Thus, unless you already know how the workload on the system is distributed over a 24-hour period, you should take measurements over the entire 24 hours.
Determining the Analysis Period Running GPA Determining the Analysis Period The object of this step is to examine the data collected with Measure in the preceding step to determine the most appropriate period of system use for GPA to analyze. Keep in mind the following: • • The analysis period should include the time of heaviest use, but should not be so narrow as to result in tuning the system for just a sharp peak.
Running GPA Running GPA Figure 2-2. Code Characters in a GPA Text Statement Paragraph Keyword Code Characters Number of lines in the paragraph :TEXT SYSTEM CPUBALANCE F 07 CPU load balancing is FAIR: Code Characters OB of CC processors on this node appear to be out of balance, considering the utilization level of this node during the measurement period. As the overall utilization of the node increases the requirement for "tighter" cpu balancing will increase. :END: CDT .
The GPA Procedure Running GPA When the MEASCOM prompt (+) appears, enter: + START MEASSUBSYS Note. You must be logged on as a super group user to invoke the START MEASSUBSYS command. You can also start the Measure subsystem in noninteractive mode by using the command: 2> MEASCOM START MEASSUBSYS After you start the Measure subsystem, log on with your normal user ID and continue the procedure with Step 3. 3. Start the Measure run as follows: a. At the TACL prompt (>), enter: 3> MEASCOM b.
The GPA Procedure Running GPA If the program executes the command successfully, continue the procedure with Step 4. If you get the following error message: 3219 MEASURE subsystem not currently running do Step 3c, below, before Step 4. (You must have super group access to perform Step 3c.) c. Start the Measure subsubsystem. Enter: + START MEASSUBSYS Delete the measfile started in Step 3b. Enter: + DELETE MEASUREMENT measfile Note.
The GPA Procedure Running GPA The system responds with status information showing the subvolume in which the active files are located. For example, the following shows that the active Measure files are in the subvolume SYS00: Process $XM00 Pri 0,50 PFR 199 P %WT 001 Userid 255,255 Program file $SYSTEM.SYS00.MEASCTL Hometerm $BMTP.#TERM2 6. Create the three structured data files GPA uses for its analysis as follows: a. At the TACL prompt, enter: 6> MEASCOM b.
The GPA Procedure Running GPA classname specifies the process class the optional report is run for. The classnames are: SYSTEM SUBSYSTEM PATHWAY SERVER TRANSIENT OTHER You can select more than one section for the optional report by setting one PARAM for each class. section-type specifies the type of output GPA should produce. The options are: SUMMARY DETAIL An optional section is run for each of the PARAMs declared.
GPA Error Messages Running GPA 11. Use PUPBAK/SCFBAK (if necessary) to restore the system to its previous condition. Enter: 18> PUP /IN PUPBAK/ or 18> SCF /IN SCFBAK/ 12. Perform other GPA tuning recommendations as appropriate (see Section 4 of the Measure User’s Guide for a general discussion of system tuning). GPA Error Messages GPA returns its error messages together with their probable causes and the suggested action to take in each case.
GPA Error Messages Running GPA Guardian Performance Analyzer (GPA) Manual— 135081 2- 10
3 Description of GPA Reports Overview of GPA Reports GPA reports include standard reports and optional reports. These reports are listed briefly below and then described in detail.
Description of GPA Reports GPA Standard Report When you specify the summary option, you receive the Process Distribution Analysis for the class. When you specify the detail option, you receive all five sections. The optional report enables you to perform several more detailed analyses than the standard report does. With it you can: • • • Identify programs or process names within a class that are consuming the most resources.
System Performance Summary Description of GPA Reports Example 3-1. Node Characteristics I. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 NODE CHARACTERISTICS CPU TYPE : CPU COUNT : TOTAL MEMORY : VOLUME COUNT : STATIC PCBS : DYNAMIC PCBS : OUT OF BALANCE CPU COUNT : MOST BUSY CPU : LEAST BUSY CPU : AVG CPU BUSY : AVG DISK VOLUME BUSY : AVG DISK PROCESS BUSY : % - RECOMMENDED RESOURCES: NODE SWAP RATE : MEASURED VALUE VLX 4 32 9 264 11 2 1 3 36.2 2.8 3.3 55.7 2.3 MAX VALUE EXPECTED 16 48 512 1024 0 0 65.
System Performance Summary Description of GPA Reports (11) AVG DISK VOLUME BUSY: The average percentage busy time of all disk volumes on the node. (12) AVG DISK PROCESS BUSY: The average of the percentage busy times of all disk processes running on the node. (13) % - RECOMMENDED RESOURCES: The extent to which the node is being utilized in relation to the maximum recommended (100 percent) level.
System Performance Summary Description of GPA Reports (12) CPU(s): The primary and backup CPUs for the volume. (13) CACHE HIT%: Percentage of requests that were satisfied with data in cache. (14) REQUEST RATE/SEC: Number of requests per second. (15) % OF ALL CACHE CALL: Cache calls for the volume as a percentage of calls for all volumes. A later section of the report, titled Disk Volume Performance Analysis, gives detailed performance data for each volume in the node.
Description of GPA Reports System Performance Summary (6) CACHE FAULT DETECTED: No cache faults detected. A cache fault occurs when a disk process expects to find a data block in cache and discovers that the memory manager has removed it. The memory manager does this whenever it needs the extra storage space to relieve a shortage of main memory. (7) OVER UTILIZED CPU: No overutilization of any processors. None of the percentage busy times for the CPUs on the node was found to be excessive.
Process and Busy Distribution Analyses Description of GPA Reports Example 3-4. System Performance Score IV. SYSTEM PERFORMANCE SCORE 1. 2. 3. 4. 5. *6. * CPU SUBSYSTEM MEMORY SUBSYSTEM DISK CACHE SUBSYSTEM DISK VOLUME SUBSYSTEM SYSTEM RECOVERY PERFORMANCE WEIGHTED ANALYSIS SYSTEM SCORE 1% 1% 83% 9% 52% 1% Average Score rated: BEST = 100, WORST = 0.
Description of GPA Reports Processor Load Balance and Performance Charts (3) SYSTEM: The SYSTEM class consists of processes that are part of the OSIMAGE or have an execution priority greater than 199. (4) SUBSYS: The SUBSYSTEM class consists of processes that are part of the system image subvolume, $SYSTEM.SYSnn.*, and are not TRANSIENT. (5) PATHWAY: The PATHWAY class consists of processes with the filename PATHMON or PATHTCP2, and are not TRANSIENT.
Processor Load Balance and Performance Charts Description of GPA Reports Example 3-6.
Description of GPA Reports Processor Load Balance and Performance Charts Processor Performance Chart This chart (Example 3-6) gives the following physical and performance data for each processor on the node analyzed: (1) CPU NUM: The processor’s identification number. (2) CPU TYPE: The processor’s type designation. All of the processors in the example are TXPs. (3) MB MEMORY: The processor’s total memory in megabytes. Each processor on this node has 8 megabytes of memory.
Disk Volume Performance Analysis Description of GPA Reports (15) CPU Count: The total number of processors. (16) Avg CPU Busy (^): The average of the percentage busy time for all processors. (17) TUNING WINDOW (w): The amount by which the percentage busy time of any processor on the node can vary from the average for all processors without being considered out of balance.
Disk Volume Performance Analysis Description of GPA Reports (6) AVG DEVICE BUSY%: The percentage of time the volume is busy, averaged for mirror devices. (7) PROCESS BUSY%: The percentage busy time for the disk I/O processes associated with the volume. (8) REQUEST RATE/SECOND: The number of requests per second. (9) CACHE HIT: The percentage of cache calls that are satisfied without a physical I/O. (10) CACHE CALLS PER REQUEST: The average number of cache calls used to satisfy a request.
Cache Performance Analysis Description of GPA Reports Cache Performance Analysis This section of the report (Example 3-9) is a table that shows cache block data for each of the disk volumes in the system. For each volume, the table gives information for each size of cache block (512 bytes, 1024 bytes, 2048 bytes, and 4096 bytes): Example 3-9.
Description of GPA Reports IPC Traffic Analysis For each processor on the node, the diagram shows: (1) CPU: The CPU’s identification number. (2) TYPE: The processor type. (3) BUSY: The percentage busy time. (4) MMGR: Pages of memory owned by the memory manager.
IPC Traffic Analysis Description of GPA Reports After the measurement is taken, enter: + LIST SERVERNET * The latter command creates a structured data file named SVNET for GPA to analyze. The IPC Traffic Analysis reports the ServerNet IPC information. This information points out which processors create (send) and which consume (receive) messages. By moving processes that create messages to the consuming CPUs, you can reduce interprocess message traffic costs. Example 3-12.
Disk Subprocess Analysis and Recommendations Description of GPA Reports (10) Write Bytes: The total number of bytes transferred as a result of write requests. (11) Write Qtime: The total time spent by write requests queued for this entity. (12) Write QBusyt: The time spent in a state in which requests for data transfer from memory to this entity were queued.
Process Move Recommendations Description of GPA Reports (1) DISK PINS: The number of disk subprocesses configured for the volume at the last SYSGEN. (2) Net Msg Rate: A measure of the workload handled by each subprocess. (See the Glossary for a more exact definition of “net messages.”) Up to eight PINs can be configured for each logical disk volume. (3) NET CHNG: The recommended change to the number of disk subprocesses for the logical volume. This change requires a SYSGEN.
Expected System Performance After Tuning Description of GPA Reports (6) The percentage busy time for the process. (This figure may be adjusted by the appropriate factor if the CPU to which the process is moved is of a different type than the original one.) (7) Number of pages in memory occupied by the process. (8) Whether the move refers to a process pair (designated by P or B) or a single process (designated by *).
GPA Analysis Statements Description of GPA Reports Example 3-15. Expected System Performance After Tuning (page 2 of 2) PROCESSOR PERFORMANCE CHART CPU CPU MB PCT DISK NUM : TYPE : MEMORY: BUSY% : PRIME : DISK CHIT MSG DISP RATE RATE RATE RATE : : : : 0 TXP 8 28.66 2 11.8 16.9 111 280.6 1 TXP 8 58.87 3 .1 197 416.9 SWAP RATE : .01 .02 MMGR PAGES : 1555 1310 PCB COUNT : 88 77 TRANSIENTS : 5 3 HALT IMPACT: MEMORY NONE 2 TXP 8 30.34 2 3 TXP 8 34.89 2 11.5 11.7 69 225.7 77 229.1 .
GPA Optional Report Description of GPA Reports GPA Optional Report When you select at least one section of the GPA Optional Report, a page at the beginning of the report summarizes your selections. Example 3-16 is an example. Example 3-16.
Summary Section Description of GPA Reports Example 3-17.
Detail Section Description of GPA Reports Detail Section Example 3-18 is an example of the additional information provided by GPA when you select the detail option for the SYSTEM class. Example 3-18.
Description of GPA Reports Detail Section (1) Busy Distribution Analysis: The Busy Distribution Analysis gives a breakdown in descending order of each process name’s CPU utilization. Note that the percentages total 99% so that the percentages are within the class, and are not for the total system. (2) Send Distribution Analysis: The Send Distribution Analysis is a breakdown by process name of the percentage of total messages sent by the class.
Description of GPA Reports Guardian Performance Analyzer (GPA) Manual— 135081 3- 24 Detail Section
4 Using GPA Information Interpreting the Standard Report A GPA analysis can detect and point to a number of causes of poor or inefficient system performance. For some of the problems, such as a memory shortage or a load imbalance, GPA can make appropriate tuning recommendations and even, in some cases, provide the means for automatically implementing the recommendations. For other performance issues, such as transient processing, GPA can reveal the condition, but the user must decide how to correct it.
System Performance Summary Using GPA Information • • The average disk volume busy time is 2.8 percent and the average disk process busy time is 3.3 percent. The system resources are being utilized at 55.7 percent of the maximum recommended level. The data indicate that this is a small- to moderate-sized system running at well below the maximum recommended level. Example 4-1. Node Characteristics for \NODEA I.
System Performance Summary Using GPA Information Example 4-3. Node Characteristics for \NODEC I. NODE CHARACTERISTICS CPU TYPE : CPU COUNT : TOTAL MEMORY : VOLUME COUNT : STATIC PCBS : DYNAMIC PCBS : OUT OF BALANCE CPU COUNT : MOST BUSY CPU : LEAST BUSY CPU : AVG CPU BUSY : AVG DISK VOLUME BUSY : AVG DISK PROCESS BUSY : % - RECOMMENDED RESOURCES: NODE SWAP RATE : MEASURED VALUE CYCLONE 4 256 28 639 5 2 2 1 58.6 11.6 2.5 69.0 .5 MAX VALUE EXPECTED 16 512 512 1024 85.
System Performance Summary Using GPA Information Figure 4-1. Response Time as a Function of CPU Utilization 8 Response 6 Time 4 Maximum recommended utilization level 0.0 CPU Utilization CDT 004 .CDD A rundown of the global performance indicators for \NODEA (Example 4-4) shows that the processor load balance on the system is poor and that there is excessive disk queuing. The cache performance, with an average cache hit percentage of 83.2, is also deficient.
System Performance Summary Using GPA Information Example 4-5. Global Performance Indicators for \NODEB III. GLOBAL PERFORMANCE INDICATORS EXCESSIVE DISPATCHING PROCESSOR LOAD BALANCE OVER UTILIZED NODE OVER UTILIZED CPU DISK VOLUME QUEUING AVERAGE CACHE HIT % : NO : FAIR : NO : NO : YES :91.38% INDEX LEVELS > 2 : OVER UTILIZED DISK : CACHE FAULT DETECTED : TRANSIENT PROCESSING : BLOCKED REQUESTS : TOTAL CACHE CALL RATE: NO NO NO NO NO 31.
System Performance Summary Using GPA Information Example 4-7. Disk Volume Performance for \NODEA II. DISK VOLUME PERFORMANCE VOLUME BUSY% REQUEST CPU CACHE REQUEST % OF ALL QTIME (s) HIT% RATE/SEC CACHE CALLS LOW BUSY DISK VOL : $X42 .0 0 2: 3 0.0% 0 0% HI BUSY DISK VOL : $SYSTEM 18.9 25 0: 1 66.1% 12 57% LOW Q-TIME VOL : $ClO 4.2 19 0: 1 90.2% 3 16% HI Q-TIME VOL : $PROJEC .0 114 2: 3 99.7% 1 4% *LOW BUSY DISK PROC: $B40 .1 0 0: 1 0.0% 0 0% *HI BUSY DISK PROC: $SYSTEM 7.2 25 0: 1 66.
Process and Busy Distribution Analyses Using GPA Information Example 4-10 shows that for \NODEC, on the other hand, the most potential for improvement is in the System Recovery Performance, the Disk Volume Subsystem, and the CPU Subsystem. Example 4-10. System Performance Indicators for \NODEC IV. SYSTEM PERFORMANCE SCORE 1. 2. 3. 4. 5. * 6.
Processor Load Balance and Performance Charts Using GPA Information Example 4-11.
Processor Load Balance and Performance Charts Using GPA Information Example 4-12.
Using GPA Information Processor Load Balance and Performance Charts could be corrected by redistributing the disk processes among the CPUs.) Note that for \NODEC (Example 4-19), the imbalance is severe enough that GPA makes a recommendation for primary disk process changes and generates “After Primary Changes” sections in its standard report. Other important parameter values shown in this chart are: • • • The page swap rate (SWAP RATE).
Processor Load Balance and Performance Charts Using GPA Information Example 4-13.
Disk Volume Performance Analysis Using GPA Information Disk Volume Performance Analysis To track the performance of the disk volume subsystem, you look at the Disk Volume Performance section (Example 4-14). Here you can see how the volumes are configured with respect to primary and backup CPUs as well as primary disk controllers. You can also see how the volumes compare with regard to a number of performance parameters.
Disk Subprocess Analysis and Recommendations Using GPA Information Example 4-15.
Processor/Disk Configuration Diagram Using GPA Information The Disk Subprocess Analysis also gives you insight into the type of disk activity that is occurring on each volume. For example, you can see that the first PIN on $DATA has a very high net message rate, yet its second PIN is only moderately high. This shows that the first PIN is able to handle many I/O requests, and that these requests, on average, are relatively inexpensive.
Processor/Disk Configuration Diagram Using GPA Information Example 4-17. Processor/Disk Configuration Diagram for \NODEA PROCESSOR/DISK CONFIGURATION DIAGRAM CPU TYPE BUSY FREE 0 TXP 34.1 8 PG 1 TXP 66.0 58 PG [CNTRLR] %01 > 1 $SYSTEM 7.2/ 824/ 13 $ClO 2.6/ 2 TNSII 32.7 513 PG 3 TNSII 11.9 1475 PG [CNTRLR] %01 > 3 $NSMS 14.7/ 1225/ 5 227/ 4 $PROJEC 6.3/ 257/ 1 $DRIVER 0.8/ 188/ 1 $XPRESS 1.7/ 295/ 1 $B40 0.1/ 0 $MEAS 0.5/ 209/ 0 $X42 0.
“After Primary Changes” Sections Using GPA Information “After Primary Changes” Sections When GPA analyzes a system, it considers whether making changes in the location of the primary disk process for each logical volume will help the overall performance of the system. If GPA determines that the performance of the system will be improved, and that there are sufficient resources available on the CPU with the backup disk process, GPA recommends changing the primary disk process to the location of the backup.
“After Primary Changes” Sections Using GPA Information Example 4-20. Processor Load Balance and Performance Charts for \NODEC After Primary Changes ESTIMATED PERFORMANCE PROFILE AFTER PRIMARY CHANGES.
“After Primary Changes” Sections Using GPA Information Example 4-21.
“After Primary Changes” Sections Using GPA Information Example 4-22. Processor/Disk Configuration Diagram for \NODEC Before Primary Changes PROCESSOR/DISK CONFIGURATION DIAGRAM CPU TYPE BUSY FREE 0 CYCLON 62.2 19576 PG 1 CYCLON 44.6 9641 PG [CNTRLR] %10 > 1 $DATA 5.4/ 1098/ 15 [CNTRLR] %32 > 2 $XL80E3 1.8/ 3962/ 16 [CNTRLR] %10 > 3 $XL80A1 3.7/ 2355/ 33 $XL80E4 2.3/ 4517/ 19 $XL80A2 2.7/ 2394/ 20 $SYSTEM 0.6/ 328/ 4 [CNTRLR] %12 > 1 $AUDIT 4.0/ 126/ 0 2 CYCLON 69.4 43764 PG 3 CYCLON 58.
Process Move Recommendations Using GPA Information Example 4-23. Processor/Disk Configuration Diagram for \NODEC After Primary Changes PROCESSOR/DISK CONFIGURATION DIAGRAM AFTER PRIMARY CHANGES CPU TYPE BUSY FREE 0 CYCLON 62.2 19576 PG 1 CYCLON 50.5 14720 PG 2 CYCLON 63.4 38685 PG 3 CYCLON 58.3 14381 PG [CNTRLR] %10 > 1 [CNTRLR] %32 > 2 [CNTRLR] %10 > 3 2<[CNTRLR] $DATA 5.4/ 1098/ 15 $XL80E3 1.8/ 3962/ 16 $XL80A1 3.7/ 2355/ 33 $V80A1 1.8/ 4082/ $SYSTEM 0.6/ 328/ $XL80E4 2.
Process Move Recommendations Using GPA Information Report Considerations No report will be produced if no candidates are found. There are no user-supplied runtime options to initiate these reports; they are STANDARD reports.
Expected System Performance After Tuning Changes Using GPA Information utilized. The CHNG CPU BUSY% is the reduction in CPU cycle consumption that will result from this switch in primary. Example 4-26.
Expected System Performance After Tuning Changes Using GPA Information Example 4-27.
Interpreting the Optional Report Using GPA Information Example 4-28.
Optional Report Section for the SYSTEM Class Using GPA Information Example 4-29.
Optional Report Section for the SYSTEM Class Using GPA Information Example 4-31 shows the additional information you receive when you specify the detail option for the SYSTEM class. From the Busy Distribution Analysis in Example 4-31, you can find what percentage of the system resources are being used on the processes in the class. Note that in the figure, the percentages add up to 100% so that the data can be described as “normalized” within the class.
Optional Report Section for the SERVER Class Using GPA Information Example 4-31.
Optional Report Section for the SERVER Class Using GPA Information The additional information in the detail section for the SERVER class is shown in Example 4-33 and Example 4-34. Program VISAOT consumes the most CPU time for the class even though it has a much lower process count than EMULD. Upon close inspection of program RCAPS510, you can see that CPU 1 has only one copy of the program while the other CPUs have two or three.
Implementing GPA Tuning Recommendations Using GPA Information Example 4-34.
The Detail System Performance Score Reports Using GPA Information corresponding tuning operations. SCFIN contains the SCF PRIMARY and SCF ALTER DISK, CACHE commands required to perform the corresponding tuning operations. If, after running PUPIN/SCFIN, you want to restore the system to its previous state, you can do so by running PUP or SCF using PUPBAK or SCFBAK as the input file. After implementing the GPA tuning recommendations, you should remeasure the system and run another GPA analysis.
Negative Contributing Factors to CPU Score Using GPA Information Example 4-35 is a cover page for the optional System Performance Score Reports. In the example, detailed sections are requested for each score. Example 4-35.
Negative Contributing Factors to Memory Subsystem Score Using GPA Information (1) Detail Negative Contributing Factors to CPU Score: Shows how the CPU subsystem score is degraded from 100% to the final resulting score due to the problems detected. These problems are represented by the following negative factors: (2) Hi Swap: CPUs that have a swap rate greater than the limit swap rate for this CPU type. A negative score gets added to the score according to how high above the limit the swap rate is.
Disk Cache Performance Score Grading Using GPA Information Disk Cache Performance Score Grading The Cache Hit Rate of each disk on the node is a positive contributing factor to the disk cache performance score. The final score of the disk cache subsystem is the average cache hit rate of all volumes. Example 4-38 is a detailed report displaying the disk cache subsystem score grading. Example 4-38.
System Recovery Performance Score Grading Using GPA Information Example 4-39. Negative Contributing Factors to Disk Volume Score (page 2 of 2) NEGATIVE CONTRIBUTING FACTORS: Blockd - Disk Blocked Problem Hiproc - High Proc Rate Indexd - Disk Reqs:Cache > 3.0 Memory - Memory Shortage Overbs Pcache Queued R:C <1 Notes on Disk Volume Subsystem Score: Good Fair Poor = = = Cpu Overbusy Problem Poor Cache Disk Queue Problem Disk reqs:Cache < 1.
CPU Failure Simulation Using GPA Information Example 4-40.
Server Process Analysis for the Individual CPUs Using GPA Information Server Process Analysis for the Individual CPUs Example 4-42 is a detailed report displaying the Server Process Analysis for CPU #00. Example 4-42.
Dynamic Server Process Analysis Using GPA Information Example 4-43 is a detailed report displaying the Server Process Analysis for CPU #04. Example 4-43.
Disk Cache Change Analysis Using GPA Information This report identifies the dynamic server processes running during the measurement, as follows: (1) Program File Name: The full file name of the dynamic server. (2) AVG Cnt: Number of dynamic servers running. (3) AVG Pri: Average priority of this server. (4) AVG Mem Pages: Average number of memory pages for this server. (5) AVG Cpu Pct: Average percentage of CPU time busy for this server. (6) AVG Disp Rate: Average dispatch rate of this server.
CPU Cache Change Analysis Using GPA Information (5) I-O Req/Sec: The disk I/O request rate per second. (6) Old Cache: The original cache pages. (7) New Cache: The new cache pages. (8) Delta Cache: The difference between the new and old cache pages. CPU Cache Change Analysis Example 4-46 is a detailed report displaying the CPU Cache Change Analysis. Example 4-46.
Error Reports Using GPA Information Error Reports The final kind of report that GPA provides is the error report. The GPA Error Report informs you of system errors and diagnoses them. There are two types of errors that can occur: • • Fatal errors that cause GPA to terminate. Non-fatal errors, on which GPA bases its performance-enhancing recommendations. Example 4-47 is an example of an error report that contains both fatal and non-fatal errors. Example 4-47.
Glossary #OUT. The GPA output file. blocked request. A request that cannot be processed because another application has blocked access to a record or file. cache. A portion of memory used to store frequently accessed information in order to save the time otherwise required for disk I/O operations. cache call. A request for disk data expected to be found in cache. cache fault. An event that occurs when a disk process expects to find a data block in cache and discovers that the memory manager has removed it.
node swap rate Glossary node swap rate. The total number of page swaps per second taking place on all the processors on a system. page swap. The transfer by a CPU of a page of data or code from disk to memory. PCB. A Process Control Block is a data area in a CPU dedicated to the control of process execution. PIN. A Process Identification Number is the numeric value that identifies a process running on a CPU. PUPBAK. A PUP command file that restores a system to its original measured state. PUPIN.
TMF Glossary TMF. The Transaction Monitoring Facility is a log-based manager that protects the integrity of a database by creating an audit trail of database changes. transient processing. CPU activity devoted to processes whose execution life is shorter than the measurement period analyzed by GPA. tuning window. The amount by which the percentage busy time for a CPU can deviate, either up or down, from the average value for the node without being considered out of balance.
tuning window Glossary Guardian Performance Analyzer (GPA) Manual— 135081 Glossary -4
Index A After primary changes 1-4, 4-9, 4-16 Average processor busy, definition 4-11 Duration, typical GPA run 1-2 Dynamic Server Process Analysis, optional report 4-37 E B Busy Distribution Analysis 3-7, 4-7 C Cache calls per request, excessive 3-12, 4-18 Cache hit rate, acceptable value 4-5, 4-12 Cache Performance Analysis 3-13, 4-12 Cache performance, relation to queue time 4-13 Cover page, optional report 3-20 CPU Cache Change Analysis, optional report 4-39 CPU Score, optional report 4-31 CPU utiliz
L Index Interrupt processing 3-8 OSIMAGE 3-20/3-21 L P Load imbalance 3-4 PARAM 3-20 PCB, static and dynamic 3-3, 3-15 Process and Busy Distribution Analyses 3-7, 4-7 Process classes, defined 3-8 Process Distribution Analysis 3-7, 4-7 Process Move Recommendations 3-17, 4-20 Process name 3-20 Processor Load Balance and Performance Charts 3-8, 4-8 Processor/Disk Configuration Diagram 3-13, 4-14 Program name 3-20 M Memory shortage cause 3-4, 4-3, 4-8 seriousness of 4-1, 4-22 Memory Subsystem Score, opt
T Index Swap rate excessive 4-3 excessive example 3-10 Swap rate, excessive 3-4 SYSGEN memory pages in output 4-10 operating system image number assigned 3-21 System Performance Summary 3-2, 4-1 System Recovery Performance Score, optional report 4-34 T Tandem analyst, when to see 4-22 Tandem products required by GPA 1-6 Transient processes, cost of 4-3 Tuning recommendations 4-29 Tuning window description 3-8, 4-11 sizing 3-11 U User benefits 1-1 Guardian Performance Analyzer (GPA) Manual— 135081 Index
U Index Guardian Performance Analyzer (GPA) Manual— 135081 Index- 4
