Tandem Performance Data Collector (TPDC) Manual Abstract The Tandem Performance Data Collector (TPDC) is an HP NonStop™ Kernel hostbased performance data collection and relationship product that significantly reduces the expertise and manpower required to collect performance data. It provides an all-inone approach to collection and integration of performance information. Product Version TPDC D20 Supported Release Version Updates (RVUs) This manual supports D48.03 and all subsequent D-series RVUs and G06.
Document History Part Number Product Version Published 102910 TPDC D20.
Tandem Performance Data Collector (TPDC) Manual Index Figures What’s New in This Manual vii Manual Information vii New and Changed Information Tables vii About This Manual ix Audience ix Related Manuals ix Manual Organization x Your Comments Invited x Notation Conventions xi 1.
2. Installing and Running TPDC Contents 2. Installing and Running TPDC Installing TPDC 2-1 Task 1: Install the TPDC Product Files 2-1 Task 2: Build a Dictionary for Enform 2-2 Task 3: Set Up SPAM 2-2 Running TPDC (Data Collection) 2-5 TPDC Output Files 2-7 Running Deferred Processing (TPDP) 2-7 3.
4. Using the System Configuration Reports (continued) Contents 4. Using the System Configuration Reports (continued) Volume Label Report 4-14 5.
5. Using the SPAM Performance Data Charts and Reports (continued) Contents 5.
Tables Contents Tables Table 2-1. Table 3-1. Table 3-2. Table 3-3. Table 3-4. Table 3-5. Table 3-6. Table 4-1. Table 5-1. Table 5-2. Table A-1.
Contents Tandem Performance Data Collector (TPDC) Manual— 526357-001 vi
What’s New in This Manual Manual Information Tandem Performance Data Collector (TPDC) Manual Abstract The Tandem Performance Data Collector (TPDC) is an HP NonStop™ Kernel hostbased performance data collection and relationship product that significantly reduces the expertise and manpower required to collect performance data. It provides an all-inone approach to collection and integration of performance information. Product Version TPDC D20 Supported Release Version Updates (RVUs) This manual supports D48.
What’s New in This Manual • • New and Changed Information Section 3, Using the Collected Data, has new entities in Table 3-2, MeasureDerived Entities, and in Table 3-3, Data Collection Entities. Appendix A, Entity and Attribute Definitions, has new entities in Table A-1, TPDC Entities. Attributes of the new entities and additional attributes in existing entities were also added.
About This Manual This manual describes how to install and use the Tandem Performance Data Collector (TPDC). TPDC is a host-based performance collection and relationship product that significantly reduces the expertise and time needed to collect performance data. TPDC integrates and normalizes data from Measure and other sources, and creates a single consolidated data file for subsequent reporting and analysis. You can use TPDC in conjunction with Enform or other HP products such as GPA and Flow Map.
Manual Organization About This Manual Manual Organization Section/Appendix Contents Section 1, TPDC Overview Provides an overview of how TPDC collects data, and describes TPDC’s functions and components. Section 2, Installing and Running TPDC Describes how to install and run TPDC, and tells about the TPDC output files. Section 3, Using the Collected Data Describes how TPDC’s components operate with data collection entities and how it generates reports.
Notation Conventions About This Manual Notation Conventions Hypertext Links Blue underline is used to indicate a hypertext link within text. By clicking a passage of text with a blue underline, you are taken to the location described. For example: This requirement is described under Backup DAM Volumes and Physical Disk Drives on page 3-2. General Syntax Notation This list summarizes the notation conventions for syntax presentation in this manual. UPPERCASE LETTERS.
General Syntax Notation About This Manual { } Braces. A group of items enclosed in braces is a list from which you are required to choose one item. The items in the list can be arranged either vertically, with aligned braces on each side of the list, or horizontally, enclosed in a pair of braces and separated by vertical lines. For example: LISTOPENS PROCESS { $appl-mgr-name } { $process-name } ALLOWSU { ON | OFF } | Vertical Line.
Notation for Messages About This Manual a blank line. This spacing distinguishes items in a continuation line from items in a vertical list of selections. For example: ALTER [ / OUT file-spec / ] LINE [ , attribute-spec ]… !i and !o. In procedure calls, the !i notation follows an input parameter (one that passes data to the called procedure); the !o notation follows an output parameter (one that returns data to the calling program).
Notation for Management Programming Interfaces About This Manual lowercase italic letters. Lowercase italic letters indicate variable items whose values are displayed or returned. For example: p-register process-name [ ] Brackets. Brackets enclose items that are sometimes, but not always, displayed.
Change Bar Notation About This Manual UPPERCASE LETTERS. Uppercase letters indicate names from definition files. Type these names exactly as shown. For example: ZCOM-TKN-SUBJ-SERV lowercase letters. Words in lowercase letters are words that are part of the notation, including Data Definition Language (DDL) keywords. For example: token-type !r. The !r notation following a token or field name indicates that the token or field is required. For example: ZCOM-TKN-OBJNAME !o. token-type ZSPI-TYP-STRING.
Change Bar Notation About This Manual Tandem Performance Data Collector (TPDC) Manual—526357-001 xvi
1 TPDC Overview The Tandem Performance Data Collector (TPDC) is an HP host-based performance data collection and relationship product for HP NonStop Kernel systems. TPDC significantly reduces the expertise and manpower required to collect performance data. It provides an all-in-one approach to collection and integration of performance information. One of the challenges in performance analysis is making sure that you have all the data necessary to perform the analysis successfully.
TPDC Components and Architecture TPDC Overview TPDC Components and Architecture TPDC has seven major components: TPDC Collects performance data and creates the consolidated performance data file. TPSFL Interrogates all disk directories on a system and records label information. TPFL Obtains label information from TPSFL data files and disk directories. TPLM Loads the consolidated data and creates Measure entities. TPLD Loads the consolidated data and creates TPDC entity files.
TPDC Overview Tandem Performance Data Collector (TPDC) Tandem Performance Data Collector (TPDC) The TPDC main component collects performance data and creates the consolidated performance data file. The unstructured consolidated performance data file contains all the data you need for subsequent analysis and reporting.
TPDC Overview Tandem Performance Load Data (TPLD) Tandem Performance Load Data (TPLD) The TPLD component loads the consolidated data file and creates component TPDC entity data files, from which you can generate optional reports using Enform and other programs. System Performance Analysis Using Measure (SPAM) The SPAM component loads the consolidated performance data (UWS) file and creates both Measure and aggregate entity files used to generate reports.
TPDC Overview How TPDC Collects Data Generating the UWS file is processor intensive. To minimize TPDC’s impact on production system performance, use TPDP (for deferred processing) to have TPDC generate reports itself. To run either of these components whenever you want after TPDC collects its data, specify the DEFER or INTERVAL options when you run TPDC. How TPDC Collects Data To serve as a useful tool throughout the performance data collection cycle, TPDC: 1.
TPDC Overview Before the Measurement Before the Measurement TPDC facilitates correct configuration of measurements by determining the hardware and software of the system to be configured and organizing the measurements based on that hardware and software configuration.
TPDC Overview After the Measurement Period After the Measurement Period After retrieving data from the Measure measurements, TPDC integrates it, normalizes it, and relates it to data from other sources. TPDC then creates a single, consolidated data (UWS) file for subsequent analysis and reporting.
TPDC Overview • • Example: Scheduling TPDC Data Collection Pathway system process activity Detailed server process activity (estimated service times and utilizations, logical I/O activity, and physical I/O activity) Example: Scheduling TPDC Data Collection This example describes a regular routine for TPDC data collection that has minimal impact on system resources: 1. At midnight, run TPSFL to create current system file information.
2 Installing and Running TPDC Topic Page Installing TPDC 2-1 Running TPDC (Data Collection) 2-5 Running Deferred Processing (TPDP) 2-7 Installing TPDC You use DSM/SCM to move the TPDC product files from the site update tape (SUT) to your disk. For more information about DSM/SCM, see the DSM/SCM User’s Guide. TPDC installation does not require a system generation (SYSGEN) and has minimal impact on the system. Task 1: Install the TPDC Product Files 1. Run DSM/SCM to install TPDC.
Installing and Running TPDC Task 2: Build a Dictionary for Enform where Componentname can be one of: TPDC TPSFL TPLM TPLD TPDP and where options represents any of the options available from the component. You can also move the component object files to $SYSTEM.SYSTEM.*. This approach lets you execute the components by only having to specify the component name (such as TPDC). TPDC, TPFL, TPSFL, and TPDP need to reside in the same volume and subvolume.
Installing and Running TPDC • Task 3: Set Up SPAM TACL routines to install and set up the package for your use and to load data from UWS files and new queries. Table 2-1. TACL Macros Included With SPAM Type of Macro Macro Name Installation TINSTALL TLOADSEG Normal operation TGATHER TSHOW THELP TPRODUCE TBUILD Task 3.1: Install SPAM With TINSTALL To install SPAM files in your current subvolume: 2> RUN TINSTALL TINSTALL uses FUP to duplicate files to your specified work subvolume.
Installing and Running TPDC Task 3: Set Up SPAM Example: SPAM Installation 1> RUN TINSTALL T8670D20 SPAM(18MAR94) TINSTALL installs your working copy of SPAM it must be run interactively from the SPAM “base directory”. It moves a subset of files from this base directory to your chosen “work directory” and creates the other files needed. The syntax for TINSTALL is: TINSTALL [/OUT /] is your chosen Tandem subvolume for your working version of SPAM (e.g. $DATA.
Installing and Running TPDC Running TPDC (Data Collection) Running TPDC (Data Collection) When you run TPDC, you can specify nine parameters: • • • Duration (in minutes) of the measurement to be taken Whether to retain the raw Measure data files after the creation of the TPDC consolidated data file Whether to generate the configuration reports, defer processing, or collect data in intervals TPDC / NAME [processname] [,OUT outfile ][,NOWAIT] / [nnnn ] [,RETAIN ] [,REPORTS ] [,NOFILE ] [,NODISCOPEN ] [,N
Installing and Running TPDC Running TPDC (Data Collection) REPORTS specifies that the listed system environment reports on the configuration of the measured system be produced.
Installing and Running TPDC TPDC Output Files DEFER enables the RETAIN option and causes TPDC to immediately terminate after the creation of the DWS file without any post-processing to create the UWS file. This option is useful to minimize processor usage by TPDC while collecting data. If you use this option, you can later run TPDP to generate a UWS file. See Running Deferred Processing (TPDP) on page 2-7.
Installing and Running TPDC Running Deferred Processing (TPDP) Using the DEFER option lets you postpone using system time to process the collected data until a less processor-intensive time on your system. You must run TPDP on the same system on which TPDC collected the data. If a UWS file is already on the system, TPDP overwrites it with the new UWS file it creates.
Installing and Running TPDC Running Deferred Processing (TPDP) the raw Measure data files and DWS file, so you can never process the data again using TPDP.
Installing and Running TPDC Running Deferred Processing (TPDP) Tandem Performance Data Collector (TPDC) Manual—526357-001 2- 10
3 Using the Collected Data Topic Page Extracting System File Label Information (TPSFL) 3-2 Creating Measure Entities for GPA and Other Programs (TPLM) 3-5 Creating TPDC Entities for Enform and Other Programs (TPLD) 3-9 Creating Performance Data Entities and Reports (SPAM) 3-12 Producing SPAM Reports 3-23 Entity Relationships 3-27 This section also contains tables of data collection entities and their relationships and lists reports that you can obtain.
Using the Collected Data Extracting System File Label Information (TPSFL) Extracting System File Label Information (TPSFL) TPDC invokes the Tandem Performance System File Label (TPSFL) program once each day to extract file label information about all disk files on a system. It stores this information in files suitable for use with all languages and query facilities available on an HP NonStop Kernel system. Figure 3-1. Tandem Performance System File Label (TPSFL) Architecture Disk Directories $volume.
Using the Collected Data Running TPSFL Prior to opening the directory on a given disk, the device is again checked for accessibility. No external sorts or programs are invoked during TPSFL execution. Each time you execute TPSFL, it creates the reports that you specify. Execution Timing Estimates On an HP NonStop S70000 system, 11 volumes and a total of 24,806 files were processed in under 60 seconds.
Using the Collected Data • TPSFL Output Files Disk report VALIDATE indicates that the subvolume specified by the subvol option is a partial subvolume name that you are using to limit the search of file labels. subvol specifies one to eight valid characters of a partial or full subvolume name. Label information is collected only from subvolumes that begin with this subvolume specification.
Using the Collected Data Creating Measure Entities for GPA and Other Programs (TPLM) Creating Measure Entities for GPA and Other Programs (TPLM) TPDC’s Load Measure (TPLM) component loads the consolidated data and creates Measure entities, from which you can generate optional reports using GPA and other programs. Table 3-2 summarizes the TPDC Measure-derived entities. For the attributes of each entity, see Appendix A, Entity and Attribute Definitions. Figure 3-2.
Using the Collected Data Entities OUT outfile names the spooler collector or existing disk file to which your output is to be directed. For a spooler collector (DEVICETYPE 0, SUBTYPE 31), level-3 spooling will be used. For a disk file, it can be entry-sequenced or an EDIT file (type 101). If you do not specify an outfile, all the output is directed to your terminal. NOWAIT returns control to your TACL now, rather than when the process completes.
Using the Collected Data Entities Table 3-2. Measure-Derived Entities Measure Entity TPDC Entity Entity Description CLUSTER CLUSTER Fiber Optic Extension (FOX) ring measurement counters and linkages. The CLUSTER entity contains information about the number of FOX (fiber optic link) messages sent and received by all processes on the measured system. CONTROLLER CONTRL Controller counters and linkages (Dxx). CPU CPU CPU measurement counters and linkages.
Using the Collected Data Entities Table 3-2. Measure-Derived Entities Measure Entity TPDC Entity Entity Description OSSNS OSSNS OSSNS counters and linkages. The OSSNS entity provides information about the operation and performance of OSS Name Server processes. PROCESS PROCESS PROCESS counters and linkages. The PROCESS entity contains information about one or more processes.
Using the Collected Data Creating TPDC Entities for Enform and Other Programs (TPLD) Creating TPDC Entities for Enform and Other Programs (TPLD) The Tandem Performance Load Data (TPLD) component loads the consolidated data file and creates component TPDC entity data files, from which you can generate optional reports using Enform and other programs. Appendix A, Entity and Attribute Definitions, shows the attributes of each entity. Figure 3-3.
Using the Collected Data Entities OUT outfile names the spooler collector or existing disk file to which your output is directed. For a spooler collector (DEVICETYPE 0, SUBTYPE 31), level-3 spooling will be used. For a disk file, it can be entry-sequenced or an EDIT file (type 101). If you do not specify an outfile, all the output is directed to your terminal. NOWAIT returns control to your TACL now, rather than when the process completes.
Using the Collected Data Entities Table 3-3. Data Collection Entities (page 2 of 2) Entity Entity Description KCCB A NonStop Kernel-derived entity that identifies the Kernel Common Control Block (KCCB) used by the Monitor and the Memory Manager. LABEL A disk directory entity that provides file label information for files accessed during measurement. LDEV A NonStop Kernel-derived entity that provides logical device information.
Using the Collected Data Creating Performance Data Entities and Reports (SPAM) Creating Performance Data Entities and Reports (SPAM) TPDC’s SPAM component uses several TACL macros to create and organize data for charts and reports: • • • • • TGATHER TSHOW TBUILD TPRODUCE THELP To initialize SPAM, use the TGATHER macro. This macro also loads the consolidated data and creates the Performance Aggregate and Measure entities that you use with Enform to generate the SPAM Performance Data Charts and Reports.
Using the Collected Data Running the SPAM TACL Macros Running the SPAM TACL Macros The TGATHER macro creates data files for your SPAM charts and reports. After you run TGATHER, you can specify another TACL macro to modify how the compiled Enform queries behave. For example, you can exclude, flag, or match various measured entities. If you did not include the TLOADSEG command in your TACLCSTM file, as explained in Task 3.2: Load the TACL Macros on page 2-3, you need to run TLOADSEG before running TGATHER.
Using the Collected Data Running the SPAM TACL Macros COLLECT LOGPROG AUTO TPRODUCE-parameters specifies automatic report production. It causes TGATHER to call TPRODUCE after a normal completion, passing the contents of TPRODUCE parameters without checking. (See the TPRODUCE macro.) Example: AUTO CSVS, PCDATA, EXCLUDE 0.01 You must specify AUTO as the last option to TGATHER. TSHOW TSHOW shows the criteria you have set to produce reports.
Using the Collected Data Running the SPAM TACL Macros When using TPRODUCE to execute any of the compiled reports, you can modify the Enform queries to exclude, flag, or match various measured entities. If you do not specify any of these options, the Enform reports use default values. You normally use TPRODUCE once or twice per data collection to run the reports. report-set specifies either: CHARTS—80-column pseudo-graphic charts of overall activity/balance.
Using the Collected Data Running the SPAM TACL Macros BEGIN 13:30, INTERVAL 5 EXCLUDE activity-exclusion-threshold enables you to set the threshold for the exclusion of low-activity entities from the output of charts and reports. The reports exclude entities whose activity is less than the threshold, letting you minimize the amount of data presented.
Using the Collected Data Running the SPAM TACL Macros Values are from 0 to 999 minutes. The default value is 0, which causes the reports to compute the interval size such that the entire window (or BEGIN time to end of window) is shown. Example: INTERVAL 5, BEGIN 08:30 IOFACTOR physIO:cache-hit-CPU-cost-ratio specifies how much more CPU cost a physical I/O (driver call minus cache hit) incurs than a cache hit.
Using the Collected Data Running the SPAM TACL Macros eight characters (possibly including spaces) that can occur anywhere within the fully qualified program file name. If you specify fewer than eight characters, the partial name is padded to the right with spaces. If your partial name includes spaces, you must enclose it in quotation marks (“).
Using the Collected Data SPAM Examples ENFORM-IN-file specifies exactly which charts, reports, or both are to be produced. The contents of the IN file should be Enform ?Execute commands to execute your chosen compiled queries. (Specify the logical name of compiled queries rather than preceding the name with an “R”.) For example: ?Execute COVER ?Execute PGOVIEW ?Execute PROFILE THELP subject THELP provides online help for any of the SPAM TACL macros.
Using the Collected Data SPAM Examples This example is the result of a TGATHER run. 1> TGATHER 940307 1035 T8670D20 SPAM(18MAR99) - COPYRIGHT COMPAQ COMPUTER CORPORATION 2000 TGATHER: Base = \CSO.$GIA03.J8670D20, Work = \CSO.$GIA03.CBSPAM; OK?(y/n)Y -----------------------------------------------------------------------TGATHER: Running \CSO.$GIA03.J8670D20.SPAM ...
Using the Collected Data SPAM Examples This example shows the usage of the TACL macro, TSHOW. 1> TSHOW T8670D20 SPAM(18MAR99) - COPYRIGHT COMPAQ COMPUTER CORPORATION 2000 -----------------------------------------------------------------------TSHOW: Checking $GIA03.CBSPAM.GLOBINFO for TPDC Load ID ...
Using the Collected Data SPAM Examples This example shows the usage of the TACL macro TPRODUCE to generate the report, COVER. 1> TPRODUCE COVER T8670D20 SPAM(18MAR99) - COPYRIGHT COMPAQ COMPUTER CORPORATION 1999 -----------------------------------------------------------------------TSHOW: Checking $GIA03.CBSPAM.GLOBINFO for TPDC Load ID ...
Using the Collected Data Producing SPAM Reports This example shows the usage of the TACL macro, THELP.
Using the Collected Data Entities TBUILD SHOW For a cross-reference of Enform reports to the data files they require, see the file XREFRPTS in the base or work subvolume. For a sample of this file, see Appendix B, Cross-Reference of Enform Queries to Data Files. Caution. The selection criteria of your queries might result in nothing selected for certain reports.
Using the Collected Data Entities Table 3-4. SPAM Performance Aggregate Entities Entity Entity Description (page 2 of 2) GLOBINFO Contains global data, including estimated free memory pages, derived from memory manager ($VIRTUAL) PROCESS entities. PGMTOTAL Contains systemwide total processes, logical file, and discopen data. PROGFILE Contains logical file data for programs, where a program is an aggregate of all processes running from the same program code file.
Using the Collected Data Entities Table 3-5.
Using the Collected Data Entity Relationships Table 3-6.
Using the Collected Data • Linkage Between Entities One is obtained from several entities. TIMETPL • One is a derived artificial name tuple. FLABNME • One is created for time management. TIME • Twenty-six are performance aggregate entities, generated by SPAM.
Using the Collected Data ASCII Name Spaces ASCII Name Spaces ASCII name spaces were projected out of all the ASCII name entities into separate entities that have duplicates removed and are in sorted order. This approach conserves space and provides more information about the use of names. TIME Fields To conserve space, all times are converted from microseconds to millisecond offsets and durations.
Using the Collected Data Tandem Performance Data Collector (TPDC) Manual—526357-001 3- 30 TIME Fields
4 Using the System Configuration Reports The TPDC System Configuration reports provide detailed information on the hardware and software configuration of the systems on which you are collecting data.
Using the System Configuration Reports Device Distribution Report Device Distribution Report The Device Distribution report summarizes the devices on the system, providing a high-level view of how the system is configured with respect to I/O devices. The possible device types are described in the System Procedure Calls Reference Manual. The Device Distribution report displays the count of each device type on the system being measured and gives a brief description of the device type.
Using the System Configuration Reports Disk Report Disk Report The Disk report gives a detailed account of the disks on the measured system, including the paths to the devices and their status. The report also details the processes controlling the devices and the space characteristics of the devices. This report is similar to the output of the PUP LISTDEV DISC command. The same codes are used for the path status.
Using the System Configuration Reports Disk Report Table 4-1. Device State Codes Code State Description blank Up/read and write The device is logically accessible to user processes. For an optical disk volume, read and write access is allowed. D Down The device is no longer logically accessible to user processes. F Format A format is in progress on a disk of subtype 16 or greater.
Using the System Configuration Reports File Label Report CH displays the backup channel connection (0 through 3). (D-series only; for G-series, GP displays the primary group number instead.) CTLR displays the backup controller number in octal. (D-series only; for G-series, MOD displays the module number instead.) UN displays the backup unit number (0 through 7). (D-series only; for G-series, SLOT displays the slot number instead.) TOT MB displays the storage capacity of disk in 1,000,000-byte increments.
Using the System Configuration Reports File Label Report VOLUME-> $TOOLS REC BLK NO PRI SEC MAX SUBVOL FILENAME END OF FILE LEN LEN EXT EXT EXT EXT ----------------- ----------- ---- ---- ---- ---- ---- ---SAFE .GUARD 327,680 990 4096 16 32 32 16 TOOLBOX .BOXBASE 376,832 4096 4096 46 4 4 256 TOOLBOX .BOXMISC 49,152 4096 4096 6 4 4 256 TOOLBOX .BOXPRS 176,128 4096 4096 22 4 4 256 TOOLBOX .BOXRLSE 40,960 4096 4096 5 4 4 256 TOOLBOX .BOXSTQ 61,440 4096 4096 8 4 4 256 TOOLBOX .
Using the System Configuration Reports Logical Device Report NO EXT displays the current number of disc extents occupied by this file. PRI EXT displays the primary extent in 2048 byte pages. SEC EXT displays the secondary extent in 2048 byte pages. MAX EXT displays the maximum number of extents this file is authorized to have. FT displays the file type of this file (0 - unstructured, 1 - entry-sequenced, 2 - relative, 3 - key-sequenced). F CDE displays the file code of this file.
Using the System Configuration Reports Logical Device -----------------------DEVICE LDEV DEV SUB NAME NUM TYPE TYPE -------- ----- ---- ---?0 0 1 0 $Z0 5 1 2 $SYSTEM 6 3 10 $TUBTS1 31 3 10 $TAPE 32 4 4 $OSP 3 6 4 $TERM0 33 6 4 $TERM1 34 6 4 $SSCP 50 13 5 $TMP 2 21 0 $IPB 58 27 0 $ZNUP 4 28 0 $ZA000 61 50 1 $ZA001 62 50 1 $LAN1 53 56 0 $SNA01 51 58 1 $SNA02 52 58 1 $NCP 1 62 0 $SUBPAS 54 63 0 $SUBNOP 55 63 0 $SUBXXX 56 63 0 $LHSUPP 57 63 5 $LXD00 59 63 3 $LXFOR 60 63 3 $LHSUP 165 63 5 Logical Device Report
Using the System Configuration Reports Logical Device Report CTLR displays the primary controller number in octal. UNIT displays the primary unit number (0 through 7). CPU displays the number of the backup CPU being measured. PIN displays the process identification number of the backup process being measured. CH displays the backup channel connection (0 through 3). CTLR displays the backup controller number in octal. UNIT displays the backup unit number (0 through 7).
Using the System Configuration Reports Node Report Node Report The Node report provides high-level characteristics of the node measured by TPDC: • • • • Name and number of the HP NonStop Kernel node Current system image Date of the system image NonStop Kernel software RVU The NODE entity provides all the information on this report. NODE ENTITY INFORMATION NODE NAME : NODE NUMBER : IMAGE NAME : SYSGEN DATE : NONSTOP KERNEL RELEASE: \CSO 240 $SYSTEM.SYS00.
Using the System Configuration Reports Process Pair Directory Report Process Pair Directory Report The Process Pair Directory report shows: • • • All named process pairs that were executing at the beginning of the measurement The creator process CPU number and process identification number (PIN) The name of the creator process if it is a named process This report identifies the named processes that existed when the measurement period began. These processes are the primary source of process creations.
Using the System Configuration Reports System Generation Values Report DCT NUM displays the destination control table number in the process pair directory. PRIMARY PID displays the primary process identification number (PID) in the process pair directory. BACKUP PPID displays the backup primary process identification number (PID) in the process pair directory. SYSTEM NAME displays the system name of the creator process. PROCESS NAME displays the process name of the creator process.
Using the System Configuration Reports Guidelines MEMORY PAGES displays the number of physical pages (two kilobytes) installed on this processor. PCBS SGEN displays the process control blocks (PCBs) generated in the measured system. LCBS SGEN displays the link control blocks (LCBs) generated in the measured system. TLES SGEN displays the number of timer list elements (TLEs) generated in the measured system. BPTS SGEN displays the number of breakpoints generated in the measured system.
Using the System Configuration Reports Volume Label Report should be in effect. On development systems, it could limit the number of concurrent debugging sessions. Volume Label Report The Volume Label report provides a one-line entry for each active logical disk volume on the system. The LABEL entity provides all the information in this report.
Using the System Configuration Reports Volume Label Report OBJ COUNT displays the number of object programs (file code 100) on this volume. EDIT COUNT displays the number of EDIT files (file code 101) on this volume. UNS COUNT displays the number of unstructured files (file code 0) on this volume. ESEQ COUNT displays the number of entry-sequenced files (file type 1) on this volume. REL COUNT displays the number of relative files (file type 2) on this volume.
Using the System Configuration Reports Tandem Performance Data Collector (TPDC) Manual— 526357-001 4- 16 Volume Label Report
5 Using the SPAM Performance Data Charts and Reports The SPAM component of the Tandem Performance Data Collector (TPDC) uses the Enform program to create the Performance Data Charts and Reports in this section. Topic Page Overview of the Charts and Reports 5-2 Performance Data Charts 5-5 Performance Data Reports 5-49 For a description of how to run SPAM and an overview of the Measure and Performance Data entities that it generates, see Section 3, Using the Collected Data.
Using the SPAM Performance Data Charts and Reports Overview of the Charts and Reports Overview of the Charts and Reports Table 5-1.
Using the SPAM Performance Data Charts and Reports Overview of the Charts and Reports Table 5-2.
Using the SPAM Performance Data Charts and Reports System Configuration System Configuration The system configuration used to generate most of the reports and charts in this section was: • • • • • • • • A D20 system. The Measure load specified a time period of 15:30:16 to 16:30:41 (3625 seconds). This system contains a single CPU type. COLLECT (Spooler Collector Program Name) during TGATHER: CSPOOL ALLOW (Process-to-File Match Tolerance) during TGATHER: 5.
Using the SPAM Performance Data Charts and Reports Performance Data Charts Performance Data Charts The charts are 80-characters wide, and the samples displayed in this section are arranged in alphabetical order. (Some extra columns were eliminated in charts in this section so they could readably fit in this format and be readable.) If you request printouts of all the charts, your output displays a table of contents for the charts. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Using the SPAM Performance Data Charts and Reports Acceleration Profiles of User Programs (AXLCHRT) Acceleration Profiles of User Programs (AXLCHRT) This 74-column chart provides an overview of accelerated versus nonaccelerated CPU consumption for non-OSIMAGE programs.
Using the SPAM Performance Data Charts and Reports Apportioned Program CPU Consumption (WKLOAD) Apportioned Program CPU Consumption (WKLOAD) This 74-column chart provides an overview of apportioned CPU consumption. The programs that consume most of the apportioned CPU time are at the top of the list.
Using the SPAM Performance Data Charts and Reports Average CPU Utilization (ACUCHRT) Average CPU Utilization (ACUCHRT) This 74-column chart, sorted by CPU number, displays the average utilization of each CPU and the component parts of this utilization including: • • • • Interrupt handling Primary disk process utilization Backup disk process utilization All other process utilization This chart uses the GLOBINFO and CPUTIL entities.
Using the SPAM Performance Data Charts and Reports • • Data Communications Line Overview (DATACOM) LID—specifies the Measure Load ID SECTION—specifies the report section number The chart is sorted in CPU number, channel, controller number, and then unit number order. Average Disc Utilization Chart for System \CSO (Release D20) -----------------------------<----Device Utilization [Average = 0.
Using the SPAM Performance Data Charts and Reports • • • • • • • Disc Cache Performance (DCPCHRT) Output bits per second Primary I/O process CPU busy percentage Primary I/O process receive queue milliseconds per request Primary I/O process CPU busy milliseconds per request Average input bytes per request Average output bytes per request Actual primary I/O process name if the line is multiple This chart uses the GLOBINFO, UNLINES, and PROCESS entities.
Using the SPAM Performance Data Charts and Reports Disc Cache Performance (DCPCHRT) The cache performance counters are updated in the Measure DISC entity by the disk process in the following order: • • Cache read: hits, misses Cache write: clean, dirty hits, misses This chart uses the DGLOBAL and DISCINFO entities.
Using the SPAM Performance Data Charts and Reports Disc File Block Splits Shares (BLKSPLT) Disc File Block Splits Shares (BLKSPLT) This 74-column chart shows which disk files are experiencing the most block splits. It is ordered by descending block splits (summed over all opens of any given file or partition), so only the files experiencing block splits are shown.
Using the SPAM Performance Data Charts and Reports • • • Disc File I/O Activity Distribution (DFIODST) A file’s contribution to a volume’s cache hits/physical I/Os Estimated CPU utilization of the primary disk process in servicing cache and I/O activity for the file Estimated utilization of the primary disk drive servicing I/O activity for the file These last two items represent what the chart calls the “Estimated Utilization Shift,” that is, by how much the CPU and primary drive utilizations would decr
Using the SPAM Performance Data Charts and Reports Disc File I/O Activity Distribution (DFIODST) page 1 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports Disc File I/O Activity Distribution (DFIODST) page 2 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports Disc File I/O Activity Distribution (DFIODST) page 3 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports Disc File I/O Activity Distribution (DFIODST) page 4 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports Disc File I/O Activity Distribution (DFIODST) page 5 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports Disc File Overview (Alphabetic) (DFOATBL) page 6 Disc File I/O Activity Distribution for System \CSO ----------------------------------- (Release D20) __________________________________________________________________________ | CPU's | | VOLUME's | | FILE's |ESTIMATED | % of | | % of | | % of |UTILIZATN | TOTAL | | CPU | | VOLUME | SHIFT Cpu Chit DIOs $Volume Chits DIOs Subvol .
Using the SPAM Performance Data Charts and Reports • • • Disc File Overview (Alphabetic) (DFOATBL) Cache hit percentage Percentage of logical requests that encountered record or file locks Average milliseconds per logical I/O This chart uses the GLOBINFO, DOPENSUM, and FILESUM entities.
Using the SPAM Performance Data Charts and Reports Disc File Overview (Alphabetic) (DFOATBL) Disc File Overview (Alphabetic) Table for System \CSO ------------------------------------Filename ======== DBCPU DBNET DBPCB DICTOBL DIRECTRY $Volume .Subvol ---------======== $LAB .NSS3 $LAB .NSS3 $LAB .NSS3 $JACK2 .SPAMD20 $GPATS2 .SYS00 $JACK .SYS00 $DATA .SYS00 $TUOMO .SYS00 $GPATS1 .SYS00 $IMAGE .SYS00 $USERS .SYS00 $WORK .SYS00 $JACK2 .SYS00 $TUBDV1 .SYS00 $DATA1 .SYS00 $DATA2 .SYS00 $DATA3 .SYS00 $DATA4 .
Using the SPAM Performance Data Charts and Reports • • • • Disc File Overview (Alphabetic) (DFOATBL) Physical I/Os per second Cache hit percentage Percentage of logical requests that encountered record or file locks Average milliseconds per logical I/O This chart uses the GLOBINFO, DOPENSUM, and FILESUM entities.
Using the SPAM Performance Data Charts and Reports Disc File Physical I/O Shares (DFPIOS) Disc File Overview (System I/Os) Table for System \CSO -------------------------------------$Volume .Subvol .Filename -------------------------$LAB .NSS3 .DBPCB $GPATS2 .SYS00 .DIRECTRY $JACK .SYS00 .DIRECTRY $DATA .SYS00 .DIRECTRY $TUOMO .SYS00 .DIRECTRY $GPATS1 .SYS00 .DIRECTRY $IMAGE .SYS00 .DIRECTRY $USERS .SYS00 .DIRECTRY $WORK .SYS00 .DIRECTRY $JACK2 .SYS00 .DIRECTRY $TUBDV1 .SYS00 .DIRECTRY $DATA1 .SYS00 .
Using the SPAM Performance Data Charts and Reports Disc File Physical I/O Shares (DFPIOS) This chart uses the GLOBINFO and DOPENSUM entities. Parameters used are: • • • LID—specifies the Measure Load ID EXCLUDE—specifies a value under which to exclude the number of requests per second SECTION—specifies the report section number The chart is sorted in descending order of total driver calls.
Using the SPAM Performance Data Charts and Reports Disc File ‘Requests Blocked’ Shares (REQBLKD) Disc File ‘Requests Blocked’ Shares (REQBLKD) This 74-column chart shows which disk files are experiencing the most requests blocked. It is ordered by descending requests blocked (summed over all opens of any given file or partition), so only the files experiencing requests blocked are shown.
Using the SPAM Performance Data Charts and Reports • • Disc I/Os per Second (DIOCHRT) Individual share of total driver calls (system I/Os) and a graph of this share Cumulative share of total driver calls This chart uses the GLOBINFO and DOPENSUM entities.
Using the SPAM Performance Data Charts and Reports • • • Disc I/Os per Second (DIOCHRT) The primary (driving) CPU The primary disk controller The drive I/O rates for writes and reads only (not seeks) This chart uses the GLOBINFO, DGLOBAL, and DISC entities. Parameters used are: • • LID—specifies the Measure Load ID SECTION—specifies the report section number The chart is sorted in CPU number, channel, controller number, and then unit number order.
Using the SPAM Performance Data Charts and Reports Disc I/Os per Second (DIOCHRT) This 74-column chart displays disk process memory utilization within each CPU. The chart includes: • • • • • CPU number Total CPU memory pages A flag if the volume is experiencing cache faults Volume name Percentage of a CPU’s memory used by the volume’s disk process, shown under either the primary or backup column but not both.
Using the SPAM Performance Data Charts and Reports Disc I/Os per Second (DIOCHRT) page 1 Disc Process Memory Utilization (by CPU) for System \CSO ---------------------------------------Memry CPU Pages === ----0 12288 <-%AGE OF-> <-CPU MEM-> $Volume Primy Bckup -------- ----- ----- <-%AGE OF TOTALS FOR ALL VOLUMES IN CPU-> <-Mem InUse-> Logicl System Cache Physcl Primry Backup Reqsts I/Os Hits I/Os ------ ------ ------ ------ ------ ------ $SYSTEM $JACK $TOOLS $TUOMO $GIA03 $WORK $DATA4 3.1% . 2.6% .
Using the SPAM Performance Data Charts and Reports Disc I/Os per Second (DIOCHRT) page 2 Disc Process Memory Utilization (by CPU) for System \CSO ---------------------------------------Memry CPU Pages === ----2 <-%AGE OF-> <-CPU MEM-> $Volume Primy Bckup -------- ----- ----- <-%AGE OF TOTALS FOR ALL VOLUMES IN CPU-> <-Mem InUse-> Logicl System Cache Physcl Primry Backup Reqsts I/Os Hits I/Os ------ ------ ------ ------ ------ ------ $MWR1 $MWR3 $GPADV2 . 0.7% . 0.7% . 0.7% ----- ----24.7% 5.2% . 13.
Using the SPAM Performance Data Charts and Reports Disc Response Time (DRTCHRT) Disc Response Time (DRTCHRT) This 80-column chart displays the average response time, in milliseconds, of each logical disk volume. This response time is the sum of the average time that a request message spends in the request queue (external to the disk process) and the average time that a request spends within the disk process (both queued and in service).
Using the SPAM Performance Data Charts and Reports Disc Response Time (DRTCHRT) Disc Response Time (Milliseconds) Chart for System \CSO --------------------------------------PC C:%PC-C:%MC $VOLUME -- ----- ----- ------02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 02 0:%32 0:%01 0:%11 0:%01-0:%11 0:%01 0:%01 0:%22 0:%21 0:%01 0:%22 0:%01 0:%01-0:%11 0:%22-0:%32 0:%21-0:%31 0:%01-0:%11 0:%11 0:%01 0:%22 0:%11 0:%21-0:%31 0:%32 0:%31 0:%31 0:%32 0:%21 $USERS $LAB $JACK2 $SYSTEM $J
Using the SPAM Performance Data Charts and Reports EXPAND Communications Line Overview (EXPAND) EXPAND Communications Line Overview (EXPAND) This 74-column chart provides a tabular overview of Expand communication lines.
Using the SPAM Performance Data Charts and Reports • • • • Logical Disc File I/O Wait Shares (LDFIOWT) The date and time of the measurement taken The system measured The numbers and types of CPUs The total, average, low, and high values for key performance metrics This chart uses the GLOBINFO entity.
Using the SPAM Performance Data Charts and Reports • • • Logical Disc File I/O Wait Shares (LDFIOWT) Timeouts and escalations Individual shares of total file busy time and a graph of this share Cumulative share of total file busy time This chart uses the GLOBINFO and FILESUM entities.
Using the SPAM Performance Data Charts and Reports SQL/MP Program Overview (SQLPROG) SQL/MP Program Overview (SQLPROG) This 76-column chart provides an overview of SQL/MP programs. It is sorted by descending elapsed busy time (summed over all SQL/MP statements executed by all processes run from each program file name) so the program objects that spent the longest time waiting for SQL/MP statement executions are at the top of the list.
Using the SPAM Performance Data Charts and Reports SQL/MP Statement Overview (SQLSTAT) SQL/MP Statement Overview (SQLSTAT) This 74-column chart identifies the longest-running SQL/MP statements. These are not individual executions of a statement, but sums of executions of the same logical statement by all running copies of programs that contain the statement. All executions of a statement by the same program-file-name, run-unit, statementindex combination are summed before this table is produced.
Using the SPAM Performance Data Charts and Reports Process Category Counts (PCOUNTS) Process Category Counts (PCOUNTS) This 74-column chart displays CPU utilization by process type. It also shows execution counts for each process type. This chart uses the GLOBINFO and CATEGORY entities.
Using the SPAM Performance Data Charts and Reports Process/CPU Priority Distribution (PRIDIST) Process/CPU Priority Distribution (PRIDIST) This 74-column chart shows utilization of each CPU at each active process priority. Owing to limited space on the page, it shows utilization in integer tenths of percent. This chart uses the GLOBINFO, CPU, and PROCESS entities.
Using the SPAM Performance Data Charts and Reports Program Disc File I/O Wait Shares (PGDIOWT) Program Disc File I/O Wait Shares (PGDIOWT) This 74-column chart provides a overview of logical disk file I/O wait times by programs, where programs are all processes running from the same program object file.
Using the SPAM Performance Data Charts and Reports Program Logical File Open Shares (PGOPENS) Program Logical File Open Shares (PGOPENS) This 74-column chart provides an overview of logical file opens by programs, where programs are all processes running from the same program object file.
Using the SPAM Performance Data Charts and Reports • • Program Overview (PGOVIEW) The program’s share of all logical file I/O wait time and a graph of this share Cumulative share of all logical file I/O wait time This chart uses the GLOBINFO, PGMTOTAL, PROGSUM, and PROGLIO entities.
Using the SPAM Performance Data Charts and Reports Program Overview (PGOVIEW) Note. This report might cover HP NonStop Kernel process pairs. It attempts to allow for this fact by considering and reporting “messages received” to be (MESSAGES-RECEIVED CHECKPOINTS). This information should be accurate for programs that use the standard NonStop Kernel Checkpointing facility, because one CHECKPOINT by the primary process equals one MESSAGE-RECEIVED by the backup process.
Using the SPAM Performance Data Charts and Reports Program Spooler File I/O Wait Shares (PGSPLWT) Program Spooler File I/O Wait Shares (PGSPLWT) This 74-column chart provides a overview of spooler file I/O wait times by programs, where programs are all processes running from the same program object file.
Using the SPAM Performance Data Charts and Reports Relative CPU Balance (RCBCHRT) Relative CPU Balance (RCBCHRT) This 74-column chart displays the relative CPU balance. The measure of balance is each CPU’s CPU-QTIME expressed as a percentage of the sum of all CPU-QTIME for all CPUs (shown as Shr% in the chart). The chart displays the component parts of this Shr%: memory queuing, other CPU queuing, and actual CPU usage. This chart uses the GLOBINFO and CPU entities.
Using the SPAM Performance Data Charts and Reports Relative Disc Balance (RDBCHRT) Relative Disc Balance (RDBCHRT) This 74-column chart displays the relative logical disk volume balance, measured by each volume’s total queuing time (external queuing plus internal queuing and service) expressed as a percentage of the sum of the queuing time of all logical volumes, shown as Shr% in the chart.
Using the SPAM Performance Data Charts and Reports TMF Overview (TMFVIEW) TMF Overview (TMFVIEW) This 80-column chart shows an overview of TMF activity in the following four categories: • • • • • • • • • • • • • • • • • Transaction rates and elapsed times Home transactions Remote transactions Transaction backouts $TMP processes CPU utilization Receive queue time Master audit-trail disk processes CPU utilization Requests per second Response time Receive queue time Master audit-trail physical disks Input
Using the SPAM Performance Data Charts and Reports TMF Overview (TMFVIEW) TMF Overview for System \DEMO ------------ (Release C30) TRANSACTION RATES/TIMES: Total ------.35 7.17 Home Transactions/Second Average Home Txn Life (sec) Remote Transactions/Second Average Remote Txn Life (sec) <----AVERAGES PER CPU-----> Overall Lowest Highest ------------------.06 .01 .16 .53 87.08 .00 2.94 Transaction BACKOUTs/Second Average Txn BACKOUT Life (sec) .00 .80 $TMP PROCESS: Average CPU Utilization Avg.
Using the SPAM Performance Data Charts and Reports Performance Data Reports Performance Data Reports The reports the printer in wide format to accommodate the 132-character lines. Due to space constraints in this manual, several of the sample reports have been truncated, with the right columns of those reports displayed below their left columns. The sample reports in this section are arranged in alphabetical order.
Using the SPAM Performance Data Charts and Reports • • • Busiest Disc Files (Logical I/Os, by Volume) (LIOBVOL) Average bytes per message sent Selectivity percentage (records used divided by records accessed) Number of disk reads per logical I/O The report aggregates all opens of the same file into one report line. All counts are in units per second. This report uses the GLOBINFO and FILESUM entities.
Using the SPAM Performance Data Charts and Reports • • • • • • • • Busiest Disc Files (Logical I/Os, by Volume) (LIOBVOL) File’s logical I/O count as percentage of volume total Average milliseconds per logical I/O Flags for lock waits, timeouts or cancels, and escalations Logical I/O counts (total and by type: read, write, and so on) Number of messages sent per logical I/O Average bytes per message sent Selectivity percentage (records used divided by records accessed) Number of disk reads per logical I/O
Using the SPAM Performance Data Charts and Reports Busiest Disc Files (Physical I/Os, by Volume) (PIOBVOL) Busiest Disc Files (Logical I/Os per Second, by Volume) Report for System \CSO -------------------------------------------------------------$Volume .Subvol .Filename ========------------------ (headings continued below) $USERS (data continued below) .SYS00 .DIRECTRY .PNA .PNAQUEUE .T9683AAB.NSSMON .T9683AAA.NSSMON .SAFE .GUARD .NSS3 .NSSLOG .#3031 .PNA .RIM .#3035 .#3033 .#3037 .T9683AAB.
Using the SPAM Performance Data Charts and Reports • • • Busiest Disc Files (System I/Os, by Volume) (SIOBVOL) LID—specifies the Measure Load ID EXCLUDE—specifies a value under which to exclude the number of requests per second SECTION—specifies the report section number The report is sorted by volume name and descending total physical I/Os. Busiest Disc Files (Physical I/Os per Second, by Volume) Report for System \CSO --------------------------------------------------------------$Volume .Subvol .
Using the SPAM Performance Data Charts and Reports Busiest Processes (by CPU) (PBSYCPU) This report uses the GLOBINFO, VSIOTOT, and DOPENSUM entities. Parameters used are: • • • LID—specifies the Measure Load ID EXCLUDE—specifies a value under which to exclude the number of requests per second SECTION—specifies the report section number The report is sorted by volume name and descending total driver calls.
Using the SPAM Performance Data Charts and Reports Busiest Processes (by CPU) (PBSYCPU) This report uses the CPU and PROCESS entities. Parameters used are: • • • • LID—specifies the Measure Load ID FLAG—specifies a percentage value under which to flag the life of a process EXCLUDE—specifies a value under which to exclude the number of requests per second SECTION—specifies the report section number The report is sorted by CPU number and descending process CPU busy time.
Using the SPAM Performance Data Charts and Reports CPU Activity (CPUREPT) CPU Activity (CPUREPT) This 132-column report shows CPU activity.
Using the SPAM Performance Data Charts and Reports CPU Activity Report for System \CSO ------------------ CPU Typ MB PCBs LCBs Static Transient === --- ---- ----- ---------- --------- CPU Activity (CPUREPT) (Release D20) <--UTILIZATION(%)--> Proc Send Total Memry Procss Intrpt Total Ovhd% Busy% ----- ----- ------ ------ ------ ------ ----- <-----PER SECOND------> Memry Proc.
Using the SPAM Performance Data Charts and Reports Disc Cache 1 (Activity, by Block Size) (CACHE1) Disc Cache 1 (Activity, by Block Size) (CACHE1) This 132-column report displays disk volume cache utilization. The four groups of columns provide the following information for each valid DP2 cache block size: • • • • • • Blocks allocated Blocks dirty Cache hits per second Cache misses per second Cache faults per second Audit buffer forces per second This report uses the DISCINFO and DISC entities.
Using the SPAM Performance Data Charts and Reports Disc Process Activity (DPREPT) Disc Process Activity (DPREPT) This 132-column report shows disk process and aggregated physical disk activity.
Using the SPAM Performance Data Charts and Reports Disc Process Activity (DPREPT) DISC PROCESS Activity Report for System \CSO ---------------------------- (Release D20) $Volume ======== === LDV PC:BC (headings continued below) ----- (1) (2) (3) <---MILLISECONDS/REQUEST---> <--------RESOURCE BUSY (%)--------> <--PER SEC--> Rqsts/ <---Discs---> Swap Free Second ExtrnlIntrnl Total Primry Backup Total Discs Primry Backup Total Primry Mirror I/O
Using the SPAM Performance Data Charts and Reports Disc Subsystem Activity (DSCREPT) DISC PROCESS Activity Report for System \CSO ---------------------------$Volume ======== $AUDIT $DATA $DATA1 LDV === 25 15 26 PC:BC ----02:03 01:02 02:03 27 8 02:03 00:01 (Release D20) (headings continued below) (data continued below) . . .
Using the SPAM Performance Data Charts and Reports Disc Subsystem Activity (DSCREPT) DISC SUBSYSTEM Activity Report for System \CSO -----------------------------$Volume LDV RQ:sec IQlen IQmsec Blkd % ======== === ------ ------ ------ -----(1) (2) (3) <--- CACHE:RQ ----> Hit Miss Flt ABF ---- ---- ---- ---(7) (8) (9) (10) (4) (5) PPID PCU ------ ---(11) (12) (Release D20) (headings continued below) (6) <--------- BUSY % --------> Total Read Write Seek ------ ------ ------ -----(13) (14) (15)
Using the SPAM Performance Data Charts and Reports EXPAND Network Line Activity (NETREPT) DISC SUBSYSTEM Activity Report for System \CSO (Release D20) -----------------------------$Volume LDV RQ:sec IQlen IQmsec Blkd % (headings continued below) ======== === ------ ------ ------ -----$AUDIT 25 .26 .00 10.8 (data continued below) $DATA 15 1.27 .00 2.9 $DATA1 26 .72 .00 3.0 $DATA2 23 .71 .00 3.7 $DATA3 19 .56 .00 3.
Using the SPAM Performance Data Charts and Reports EXPAND Network Line Activity (NETREPT) report contains separate report lines for the path handler and for each line handler. Not all counter values are reported for each of these entity types. This report uses the GLOBINFO and EXLINE entities. Parameters used are: • • LID—specifies the Measure Load ID SECTION—specifies the report section number The report is sorted by device name.
Using the SPAM Performance Data Charts and Reports Global Indicators (COVER) Global Indicators (COVER) This 74-column report is a cover page for the set of charts and reports. It shows global information, which includes: • • • • The date and time of the measurement taken The system measured The numbers and types of CPUs The total, average, low, and high values for key performance metrics This report uses the GLOBINFO entity.
Using the SPAM Performance Data Charts and Reports Low Selectivity (Needle in a Haystack) SQL Table Access Report (HAYSTCK) Low Selectivity (Needle in a Haystack) SQL Table Access Report (HAYSTCK) This 132-column report shows specific opens of SQL tables that exhibit low selectivity—where the ratio of records used to records accessed is relatively small.
Using the SPAM Performance Data Charts and Reports Low Selectivity (Needle in a Haystack) SQL Table Access Report (HAYSTCK) Low Selectivity (Needle in a Haystack) SQL Table Access Report for System \P20ET -------------------------------------------------------------- (C30) Sec <--WINDOW--> <---OPEN---> per <----TABLE NAME-------> [HH:MM]Select- Records Disc Records Disc <-------OPENING PROCESS-------> Log Filename $Volume.
Using the SPAM Performance Data Charts and Reports Network System Activity (SYSREPT) Low Selectivity (Needle in a Haystack) SQL Table Access Report for System \P20ET (C30) -------------------------------------------------------------- <-------TABLE NAME-------> Filename $Volume .
Using the SPAM Performance Data Charts and Reports Network System Activity (SYSREPT) Parameters used are: • • LID—specifies the Measure Load ID SECTION—specifies the report section number The report is sorted by originating system name and destination system name.
Using the SPAM Performance Data Charts and Reports New Process Creation Timeline (NEWPROC) New Process Creation Timeline (NEWPROC) This 132-column report shows new process creations in each of 16 time intervals. The report displays: • • • Program file name Total number of executions found Number of processes created some time in each interval This report uses the GLOBINFO and PROCESS entities.
Using the SPAM Performance Data Charts and Reports • • • • • Physical Disc I/O Activity (PDISCIO) Messages received per second Messages sent per second Ratio of messages sent to messages received Average CPU milliseconds consumed per message received Average CPU milliseconds consumed per message sent This report uses the CPU and PROCESS entities.
Using the SPAM Performance Data Charts and Reports • • • • Process Existence Timeline (PTIMLIN) I/Os per seconds, of each type Bytes transferred per second, in and out Average bytes per I/O, in and out Average milliseconds per physical I/O, in and out This report uses the DGLOBAL, DISCINFO, and DISC entities.
Using the SPAM Performance Data Charts and Reports • • Process Memory Utilization (by CPU) (PMEMCPU) Total number of executions found Number of processes existing for at least part of each interval This report uses the GLOBINFO and PROCESS entities.
Using the SPAM Performance Data Charts and Reports Process Memory Utilization (by CPU) (PMEMCPU) This report uses the CPU and PROCESS entities. Parameters used are: • • • • LID—specifies the Measure Load ID FLAG—specifies a percentage value under which to flag the life of a process EXCLUDE—specifies a value under which to exclude the number of requests per second SECTION—specifies the report section number The report is sorted by CPU number and descending present pages queue time.
Using the SPAM Performance Data Charts and Reports Process Page Faults (by CPU) (PFLTCPU) Process Page Faults (by CPU) (PFLTCPU) This 132-column report shows processes in each CPU that suffered the most from page fault activity.
Using the SPAM Performance Data Charts and Reports Server Summary (from $RECEIVE File) (SERVICE) Process Page Fault Activity (by CPU) Report for System \P20ET ------------------------------------------- (C30) Process PID -------- ==---- Pri --- Program File Name -------------------------- (headings continued below) $W012 139 135 150 139 $SYSTEM $USER01 $SYSTEM $SYSTEM (data continued below) 0,172 0,191 0,174 0,194 .SYSTEM .OPAXOBJ .SYSTEM .SYSTEM .SQLCI .OPAS012A .
Using the SPAM Performance Data Charts and Reports Server Summary (from $RECEIVE File) (SERVICE) Parameters used are: • • LID—specifies the Measure Load ID SECTION—specifies the report section number The report is sorted by process.
Using the SPAM Performance Data Charts and Reports Server Summary (from $RECEIVE File) (SERVICE) Server Summary (from $RECEIVE File) Report for System \P20ET ------------------------------------------ (C30) <-ESTIMATED--> Srvice Utili- Fil <-[HH:MM]-> Elaps Trans/ Av RCV Time zation Opn Procss PID Pri Program File Name From To Scnds Second Q msec (msec) (pct) Cnt ====== ----- --- ----------------======== ===== ------ ----- --------- ------ ------- --$DHC2 $DHC3 $DHC4 $DHCF $A088 $B088 1,143 2,145 3,14
Using the SPAM Performance Data Charts and Reports TMF Activity Report (TMFREPT) TMF Activity Report (TMFREPT) This 132-column shows the TMF activity in each CPU over the measurement interval. The report includes: • • • The number of transactions per second The average and maximum number of transactions The average number of seconds that it took a transaction to process This report uses the TMF entity.
Using the SPAM Performance Data Charts and Reports TMF Activity Report (TMFREPT) Tandem Performance Data Collector (TPDC) Manual— 526357-001 5- 80
A Entity and Attribute Definitions Table A-1 gives an alphabetical overview of all the entities used by TPDC and indicates the entity type as used by TPDC. Following this table, each entity is listed and defined by its attribute identification. For more information, see the DDL records in your distribution subvolume. Table A-1. TPDC Entities (page 1 of 5) Entity TPDC Entity Type ALTK Data Collection A disk directory entity that provides information about alternate key files.
Entity and Attribute Definitions Table A-1. TPDC Entities (page 2 of 5) Entity TPDC Entity Type DISCOPEN Measure-Derived DISCOPEN counters and linkages. The DISCOPEN entity contains information about which programs are accessing which files, including program usage and process activity (physical file access). DISK Measure-Derived Derived from the Measure DISC entity; contains information about disk counters and linkages. DISKFILE Measure-Derived DISKFILE summary counters and linkages (D20).
Entity and Attribute Definitions Table A-1. TPDC Entities (page 3 of 5) Entity TPDC Entity Type LDEVNUM Data Collection A NonStop Kernel-derived entity that provides logical device number information. LDISC Measure-Derived Derived from the Measure DISC entity, contains information about logical disk counters and linkages. LINE Measure-Derived LINE counters and linkages. The LINE entity contains information about one or more communications lines.
Entity and Attribute Definitions Table A-1. TPDC Entities (page 4 of 5) Entity TPDC Entity Type PPD Data Collection A NonStop Kernel-derived entity that identifies process pair directory (PPD) named processes running a process pair. PROCESS Data Collection PROCESS counters and linkages. The PROCESS entity contains information about one or more processes. PROCNME Data Collection An ASCII name entity that identifies processes.
Entity and Attribute Definitions Table A-1. TPDC Entities (page 5 of 5) Entity TPDC Entity Type TERMINAL Data Collection TERMINAL counters and linkages. The TERMINAL entity contains information about the terminal I/O. TIME Data Collection A time management entity that gives execution information and LCT and GMT base times. TIMETPL Data Collection A time management entity that gives a single ordered timeline for all entities measured.
Entity and Attribute Definitions ALTK (continued) system-number System number keytag-value Key tag value altkey-offset Alternate key offset in record altkey-length Alternate key length in bytes altkey-null-value Alternate key null value for no updates null-altk Flag for null values unique-altk Flag for unique alternate keys update-altk Flag for update status insertionorder-altk Flag for insertion order CACHE Attribute Identification Definition cache-link Identification field for CACHE
Entity and Attribute Definitions CATEGORY (continued) pdesc Process description pcount Process count pbusy-time Process busy time CLUSTER Attribute Identification Definition cluster-link Identification field for CLUSTER messages-sent Measure field - messages sent sent-bytes Measure field - bytes sent returned-bytes Measure field - bytes returned messages-received Measure field - messages received received-bytes Measure field - bytes received reply-bytes Measure field - reply bytes nod
Entity and Attribute Definitions CONTROL (continued) ctrl Measure field - controller ctrl-type Measure field - controller type io-qlen-max Measure field - maximum I/O queue length CPU Attribute Identification Definition cpu-link Identification field for CPU memory-pages Measure field - memory pages cpu-qtime Measure field - CPU queue time mem-qtime Measure field - memory queue time cpu-busy-time Measure field - CPU busy time intr-busy-time Measure field - interrupt busy time process-ovh
Entity and Attribute Definitions CPU (continued) starting-free-mem Measure field - free frames at the start of measurement ending-free-mem Measure field - free frames at the end of measurement starting-ucme Measure field - UCME frames at the start of measurement ending-ucme Measure field - UCME frames at the end of measurement starting-uds.
Entity and Attribute Definitions CPUX Attribute Identification Definition cpux-sequence Identification field for CPUX loadid Identification field for LOAD load-id Redefines LOAD prefix-id Identification field for load prefix interval-id Identification field for load interval cpu-num CPU number from-timestamp Measure field - from time to timestamp Measure field - to time delta-time Measure field - elapsed time system-name Measure field - name of system os-version Measure field - opera
Entity and Attribute Definitions CPUX (continued) tns-busy-time Measure field - transaction busy time comp-traps Measure field - compatibility traps static-pcbs SPAM field for static process control blocks with a life greater than or equal to 90% of CPUs transient-pcbs SPAM field for transient process control blocks with a life less than 90% of CPUs os-cpu-busy-time SPAM field for operating system CPU busy time os-messages SPAM field for operating system messages sent and received total-message
Entity and Attribute Definitions DEVICE (continued) write-qlen-max Measure field - maximum number of write requests queued device-qbusy-time Measure field - busy time if any kind of request is queued input-bytes-f Measure field - 64-bit version of input bytes output-bytes-f Measure field - 64-bit version of output bytes scsi-id Measure field - SCSI port identifier dbio-reads Measure field - read operation enabled for direct bulk I/O dbio-writes Measure field - write operation enabled for direc
Entity and Attribute Definitions DGLOBAL (continued) hi-cache-hits High cache hits total-cache-misses Total cache misses lo-cache-misses Low cache misses hi-cache-misses High cache misses total-cache-faults Total cache faults lo-cach-faults Low cache faults hi-cache-faults High cache faults total-cache-forces Total cache forces lo-cache-forces Low cache forces hi-cache-forces High cache forces total-swaps Total number of page swaps lo-swaps Low number of page swaps hi-swaps High nu
Entity and Attribute Definitions DGLOBAL (continued) lo-pages-qtime Low memory pages queue time hi-pages-qtime High memory pages queue time total-disc-busy-time Actual disk busy time adj-total-busy-time Busy times adjusted for seeks/mirroring DISCINFO Attribute Identification Definition discinfo-sequence Identification field for DISCINFO loadid Identification field for LOAD load-id Redefines LOADID prefix-id Identification for load prefix interval-id Identification for load interval sys
Entity and Attribute Definitions DISCINFO (continued) swaps Number of page swaps frees Free space reqs-blkd Requests blocked total-messages Total messages p-messages Primary messages b-messages Backup messages total-qtime Total queue and service time req-time Internal queue and service time recv-qtime External queue and service time total-ready-time Total DP ready time p-ready-time Primary DP ready time b-ready-time Backup DP ready time total-cpu-busy-time Total DP CPU busy time p
Entity and Attribute Definitions DISCINFO (continued) mirror-counts Counts in mirror drive m-reads Reads in mirror drive m-writes Writes in mirror drive m-input-bytes Input bytes in mirror drive m-output-bytes Output bytes in mirror drive m-read-busy-time Mirror read busy time, including apportioned seeks m-write busy-time Mirror write busy time, including apportioned seeks DISCOPEN Attribute Identification Definition discopen-link Identification field for DISCOPEN lockwait-time Measure
Entity and Attribute Definitions DISCOPEN (continued) dbio-output-calls Measure field - number of direct bulk I/O write operations sql-operation-time Measure field - DP2 process busy time working on SQL MID Measure-field - unique identifier for OSS files pathid Measure field - internal representation of OSS file crvsn Measure field - creation volume sequence number DISK Attribute Identification Definition disk-link Identification field for DISK cap Capacity in 2048-byte units free Beginnin
Entity and Attribute Definitions DISK (continued) mirrored Mirrored flag (0/1) pri-unit Primary unit number mir-unit Mirror unit number disk-path Disk path array (8) entries current-path Current path flag (*) inactive valid-path Valid path flag (0/1) channel Channel address of device controller Controller address of device dev-unit Device unit number dev-state Device state (See PUP Reference Manual.
Entity and Attribute Definitions DISKFILE Attribute Identification Definition diskfile-link Identification field for DISKFILE lockwait-time Measure field-lock wait time max-lockwait-time Measure field- maximum lock wait time open-qtime Measure field-file open queue time starting-eof Measure field- starting end-of-file ending-eof Measure field- ending end-of-file transient-opens Measure field-transient opens driver-input-calls Measure field- driver input calls driver-output-calls Measure
Entity and Attribute Definitions DISKFILE (continued) crvsn Measure field - creation volume sequence number oss-block-writes Measure field - block writes issued by DP2 for OSS file manager oss-block-write-bytes Measure field - bytes written by DP2 for OSS file manager oss-cache-callbacks Measure field- cache callbacks sent by DP2 to the OSS file manager oss-callback-writes Measure field - OSS-BLOCK-WRITES received by DP2 DPENSUM Attribute Identification Definition dopensum Identification field
Entity and Attribute Definitions DPENSUM (continued) total-counts Total physical I/O counts opens Open counts driver-counts Driver counts drtotal Sum of next two counters drins Driver input counts drouts Driver output counts cache-counts Cache counts chtotal Sum of next two counters crhits Cache read hits cwhits Cache write hits splits Block splits stalls Stalls physicals Driver counts plus cache counts EXLINE Attribute Identification Definition exline-sequence Identification f
Entity and Attribute Definitions EXLINE (continued) reads Number of reads writes Number of writes 12in-bytes Level-2 input bytes 12out-bytes Level-2 output bytes din4-bytes Level-4 input data bytes dout4-bytes Level-4 output bytes cin4-bytes Level-4 input bytes cout4-bytes Level-4 output bytes u64-bytes Unit 64 bytes u128-bytes Unit 128 bytes u256-bytes Unit 256 bytes u512-bytes Unit 512 bytes u1024-bytes Unit 1024 bytes u2048-bytes Unit 2048 bytes u4096-bytes Unit 4096 bytes
Entity and Attribute Definitions FILE (continued) opennme-link Identification for OPENNME openproc-link Identification for OPENPROC process-link Identification for PROCESS timetpl-link Identification for TIMETPL device-type Measure field- device type file-type Measure field-file type message-bytes-f Measure field - 64-bit version of message-bytes device-name Measure field - disk device name opener-device-name Measure field - disk device name of opener program dbio-reads Measure field - d
Entity and Attribute Definitions FILESUM (continued) volume Volume name partial-name Partial name for subvolume and file name subvol Subvolume name filename File name opens File opens logical-counts Logical I/O counts by all openers of the file busy-time Busy time I/O counts by file openers reads Read I/O counts by file openers writes Write I/O counts by file openers updates Update I/O counts by file openers deletes Delete I/O counts by file openers infos Information I/O counts by fi
Entity and Attribute Definitions GLOBINFO Attribute Identification Definition globinfo-sequence Identification field for GLOBINFO loadid Identification field for LOAD load-id Redefines LOAD prefix-id Identification field for PREFIX interval-id Identification field for INTERVAL system-name Name of system os-version Operating system version letter Operating system letter number Operating system number from-timestamp From time to-timestamp To time delta-time Elapsed time hybrid-syste
Entity and Attribute Definitions GLOBINFO (continued) lo-cpu-qtime Low CPU queue time hi-cpu-qtime High CPU queue time total-mem-qtime Total memory queue time lo-mem-qtime Low memory queue time hi-mem-qtime High memory queue time total-mem-pages Total memory pages lo-mem-pages Low memory pages hi-mem-pages High memory pages total-memman-pages Total memory manager pages lo-memman-pages Low memory manager pages hi-memman-pages High memory manager pages total-est-free-pages Total estima
Entity and Attribute Definitions GLOBINFO (continued) lo-response-time Low response time hi-response-time High response time total-static-pcbs Total static process control blocks with a life greater than or equal to 90% of CPUs lo-static-pcbs Low static process control blocks with a life greater than or equal to 90% of CPUs hi-static-pcbs High static process control blocks with a life greater than or equal to 90% of CPUs total-transient-pcbs Total transient process control blocks with a life les
Entity and Attribute Definitions GNAME Attribute Identification Definition gname-link Identification field for GNAME gname-name NonStop Kernel system process name gname-occurs Number of times name used KCCB Attribute Identification Definition cpu-num CPU number of data source sample-id Sample identification number delta-base-time Milliseconds from measurement start cme-events Correctable memory errors iocycles Number of cycles for I/O pages clock-calls Number of times a page frame sear
Entity and Attribute Definitions KCCB (continued) dispatch-ints Dispatch interrupts other-ints All other interrupts msgs-successful Successful messages msgs-unsuccessful Unsuccessful messages msgs-stolen Buffer management cpu-busy-pct Percentage of CPU busy time cpu-idle-pct Percentage of CPU idle time cpu-intr-pct Percentage of CPU interrupt time freepages Pages of memory available currlocks Pages of memory currently locked iocurr Current I/O pages in use iofails Failures to allocat
Entity and Attribute Definitions LABEL Attribute Identification Definition label-link Identification field for LABEL flabnme-link Identification field for FLABNME eof Byte current end of file filecode HP file code filetype HP file type rec-length Record length block-length Block length pri-extent Primary extent in 2048-byte pages sec-extent Secondary extent in 2048-byte pages current-extents Number of extents in use max-extents Maximum number of extents allowed key-length Key lengt
Entity and Attribute Definitions LABEL (continued) create-gmt-timestamp File creation timestamp in GMT modify-timestamp File last modify timestamp lastop-timestamp File last open timestamp LDEV Attribute Identification Definition ldev-link Identification field for LDEV ldevnme-link Identification field for LDEVNME disk-link Identification field for DISK device-number Logical device number device-type Device type sub Device subtype pcpu Primary CPU number pchn Primary channel number
Entity and Attribute Definitions LDEVNME Attribute Identification Definition ldevnme-link Identification field for LDEVNME ldev-link Identification field for LDEV ldevnme-nme Logical device name device-name Synonym for ldevnme-nme exclusion TPDC application field category TPDC application field LDEVNUM Attribute Identification Definition LDEVNUM-LINK Identification field for LDEVNUM ldevnme-link Identification field for LDEVNME ldev-link Identification field for LDEV ldevnum-number L
Entity and Attribute Definitions LINE Attribute Identification Definition line-link Identification field for LINE read-busy-time Measure field - read busy time write-busy-time Measure-field - write bust time input-bytes Measure field - number of input bytes input-data-bytes Measure field - number of input data bytes output-bytes Measure field - number of output bytes output-data-bytes Measure field - number of output data bytes reads Measure field - number of reads requests Measure field
Entity and Attribute Definitions MACHINE (continued) hist-0513-1024 Messages 513-1024 bytes hist-1025-2048 Messages 1025-2048 bytes hist-2049-4096 Messages 2048-4096 bytes hist-gt-4096 Messages greater than 4096 bytes hist-page-faults Page faults interrupts hist-bus-recieve Bus receive interrupts hist-timer-lists Timer lists interrupts hist-std-io Standard I/O interrupts hist-dispatch Dispatch interrupts hist-other Other interrupts hist-successful Successful expedited messages hist-un
Entity and Attribute Definitions MACHINE (continued) mavg-successful Average measurement successful expedited messages mavg-unsuccessful Average measurement unsuccessful expedited messages mavg-stolen Average measurement stolen expedited messages mavg-busy-pst Average measurement percentage CPU busy time mavg-idle-pct Average measurement percentage interrupt time mavg-intr-pct Average measurement percentage idle time mavg-highlocks Average measurement highest lock memory page count mavg-max-s
Entity and Attribute Definitions NETLINE Attribute Identification Definition netline-link Identification field for NETLINE read-busy-time Measure field - read busy time write-busy-time Measure field - write busy time cin4-bytes Measure field - level-4 input controller bytes cout4-bytes Measure field - level-4 output controller bytes din4-bytes Measure field - level-4 input data bytes dout4-bytes Measure field - level-4 output data bytes 12in-bytes Measure field - level-2 input bytes 12out
Entity and Attribute Definitions NODE Attribute Identification Definition last-sysproc-entry Last system procedure entry sysgen-date Operating system image create date osi-year Operating system image create year osi-month Operating system image create month osi-day Operating system image create day osi-hours Operating system image create hours osi-minutes Operating system image create minutes osi-seconds Operating system image create seconds node-number Node number of system device-num
Entity and Attribute Definitions NODENUM Attribute Identification Definition nodenum-link Identification field for NODENUM nodenme-link Identification field for NODENME nodenum-number Node number system-number Synonym for node number OPDISK Attribute Identification Definition opdisk-link Identification field for OPDISK read-busy-time Measure field- read busy time request-qtime Measure field- request queue time seek-busy-time Measure field- seek busy time write-busy-time Measure field-
Entity and Attribute Definitions OSSCPU (continued) pp-block-size Measure field - size of an OSS pipe pool block in bytes pp-blocks-allowed Measure field - maximum blocks allowed for the OSS pipe pool pxs-ending-bytes Measure field - current bytes allocated within the POSIX extended segment pxs-failures Measure field - allocation failures for POSIX extended segment fs-direct-reads Measure field - OSS file-system noncached read requests fs-direct-read-bytes Measure field - total bytes read by OSS
Entity and Attribute Definitions OSSCPU (continued) fm-callbacks Measure field - cache callbacks received by the OSS file manager fm-callback-wt-reqs Measure field - cache writes issued by the OSS file manager fm-callback-wt-blocks Measure field - blocks written back by the OSS file manager fm-stolen-blk-wt-reqs Measure field - times OSS cache block writes to clean pp-block-inuse-qtime Measure field - time the OSS pipe pool blocks were in use pp-block-inuse-qmax Measure field - max blocks in use
Entity and Attribute Definitions OSSNAMES file-system-name Measure field - system name where the file resides file-name Measure field - Guardian fie name volume Measure field - volume name subvol Measure field - subvolume name filename Measure field - file name pathname-len Measure field - length of pathname pathname Measure field - OSS pathname OSSNS Attribute Identification Definition ossns-link Identification field for OSSNS timetpl-link Identification field for TIMETPL cpu-num CPU
Entity and Attribute Definitions OSSNS (continued) dp2-dd-reqs Measure field - data definition requests sent to DP2 dp2-messages Measure field - messages sent to DP2 dp2-msg-qtime Measure field - time spent in a thread waiting dp2-msg-qlen-max Measure field - max number of DP2 messages outstanding ps-messages Measure field - messages to pipe server processes ps-msg-qtime Measure field - time, pipe server messages were outstanding ps-msg-qlen-max Measure field - max concurrent pipe server messa
Entity and Attribute Definitions PDISC logical-device Measure field- logical device unit-number Measure field- unit number timetpl-link Identification field for TIMEPL read-qbusy-time Measure field - busy time if read requests are in queue read-qtime Measure field - total time spent by read requests in queue read-qlen-max Measure field - max number of read requests queued write-qbusy-time Measure field - busy time if write requests are in queue write-qtime Measure field - total time spent by
Entity and Attribute Definitions PGMTOTAL (continued) sent-bytes Number of bytes sent returned-bytes Number of bytes returned messages-received Number of messages received received-bytes Number of bytes received reply-bytes Number of reply bytes lcbs-inuse-qtime Amount of queue time for link control blocks in use checkpoints Number of checkpoints accel-busy-time TNS/R accelerator busy time tns-busy-time TNS/R busy time comp-traps TNS/R compatibility traps program-accelerated TNS/R prog
Entity and Attribute Definitions PGMTOTAL (continued) spl-logical-ios Number of spooler file logical I/Os spl-messages Number of spooler file messages spl-message-bytes Number of spooler file message bytes logical-process-filetotals Logical process file counters follow: lpf-opens Number of logical process file opens lpf-file-busy-time Logical process file busy time lpf-logical-ios Number of logical process file logical I/Os lpf-messages Number of logical process file messages lpf-message-by
Entity and Attribute Definitions PPD Attribute Identification Definition ppd-link Identification field for PPD nodenme-link Identification field for NODENME process-name Eight-character name beginning with “$” dct-name Destination control table number pcpu Primary CPU number ppin Primary PIN bcpu Backup CPU number bpin Backup PIN crsys Creator system ID number crcpu Creator CPU number crpin Creator PIN crname Creator name PROCESS Attribute Identification Definition process-link
Entity and Attribute Definitions PROCESS (continued) reply-bytes Measure field- number of bytes replied returned-bytes Measure field- number of returned bytes sent-bytes Measure field- number of sent bytes vsems Measure field- number of times the process waited for a resource in use by another process, where the resource is protected by the NonStop Kernel semaphore facility ancpid-link Identification field for ancestor PIDNUM ancproc-link Identification field for ancestor PROCNME ancsys-link I
Entity and Attribute Definitions PROCESS (continued) sent-bytes-f Measure field - 64-bit version of sent-bytes returned-bytes-f Measure field - 64-bit version of returned-bytes received-bytes-f Measure field - 64-bit version of received-bytes reply-bytes-f Measure field - 64-bit version of reply-bytes pres-pages-start Measure field - pages sponsored by process at start pres-pages-end Measure field - pages sponsored by process at end abs-segs-qtime Measure field - time that segment table entrie
Entity and Attribute Definitions PROCESS (continued) launches Measure field - creations of other processes launch-wait-time Measure field - wait time on launch requests open-close-wait-time Measure field - wait time on the open and close requests PROCNME Attribute Identification Definition procnme-link Identification field for PROCNME procnme-nme Eight-character process name, starting with $ procnme-occurs Number of times name was used process-link Identification field for PROCESS exclusion
Entity and Attribute Definitions PROGFILE (continued) opens Number of logical file opens file-busy-time Amount of logical file busy time logical-ios Number of logical file I/Os input-logical-ios Number of logical file input I/Os output-logical-ios Number of logical file output I/Os reads Number of logical file reads writes Number of logical file writes updates-or-replies Number of logical file updates or replies info-calls Number of logical file deletes or write-reads records-used Number
Entity and Attribute Definitions PROGLIO (continued) tot-output-logical-ios Total output logical I/Os tot-reads Total logical reads tot-writes Total logical writes tot-updates-or-replies Total logical updates or replies tot-deletes-or-writereads Total logical deletes or writereads tot-info-calls Total logical information calls tot-messages Total logical messages tot-message-bytes Total logical message bytes tot-timeouts-or-cancels Total logical timeouts or cancels logical-disc-file-counts
Entity and Attribute Definitions PROGLIO (continued) lpf-file-busy-time Logical process file busy time lpf-logical-ios Number of logical process file logical I/Os lpf-reads Number of logical process file reads lpf-writes Number of logical process file writes lpf-writereads Number of logical process writereads lpf-info-calls Number of logical process file information calls lpf-messages Number of logical process file messages lpf-message-bytes Number of logical process file message bytes lpf-
Entity and Attribute Definitions PROGSIO (continued) load-id Redefines LOADID prefix-id Identification field for load prefix interval-id Identification field for load interval program-file-name Name of program file volume Program file volume subvol Program file subvolume filename Program file file name total-discopens Total disk opens dio-counts Driver I/O counts follow: total-driver-calls Total driver calls driver-input-calls Number of driver input calls total-cache-hits Total cache
Entity and Attribute Definitions PROGSQL (continued) disc-reads Number of disk reads messages Number of messages message-bytes Number of message bytes sorts Number of sorts elapsed-sort-time Amount of elapsed sort time recompiles Number of recompiles elapsed-recompile-time Amount of elapsed recompile time lock-waits Number of lock waits timeouts Number of timeouts escalations Number of escalations PROGSUM Attribute Identification Definition progsum-sequence Identification field for P
Entity and Attribute Definitions PROGSUM (continued) recv-qtime Amount of receive queue time messages-sent Number of messages sent sent-bytes Number of bytes sent returned-bytes Number of returned bytes messages-received Number of messages received received-bytes Number of received bytes reply-bytes Number of reply bytes lcbs-inuse-qtime Number of link control blocks in use queue time checkpoints Number of checkpoints accel-busy-time TNS/R accelerator busy time tns-busy-time TNS/R tran
Entity and Attribute Definitions PROGSUM (continued) spooler-file-totals Spooler file counters follow: spl-opens Number of spooler file opens spl-file-busy-time Spooler file busy time spl-logical-ios Number of spooler file logical I/Os spl-messages Number of spooler file messages spl-messages-bytes Number of spooler file message bytes logical-process-filetotals Logical process file counters follow: lpf-opens Number of logical process file opens lpf-file-busy-time Logical process file busy
Entity and Attribute Definitions SERVER Attribute Identification Definition loadid Identification filed for LOAD load-id Redefines LOADID prefix-id Identification field for load prefix interval-id Identification field for load interval program-file-name Program file name volume Program volume subvol Program subvolume name filename Program file name process-name Process name from-timestamp From time to-timestamp To time delta-time Elapsed time pid Group level name for process ID d
Entity and Attribute Definitions SQLPROC (continued) sql-time-newprocess-time Measure field- SQL new process time ancpid-link Identification field for ancestor PIDNUM ancproc-link Identification field for PROCNME ancsys-link Identification field for ancestor system creator-link Identification field for creator USERNUM flabnme-link Identification field for FLABNME procnme-link Identification field for PROCNME process-link Identification PROCESS timetpl-link Identification field for TIMETPL
Entity and Attribute Definitions SQLSTMT (continued) creator-link Identification field for creator USERNUM flabnme-link Identification field for FLABNME process-link Identification field for PROCESS procnme-link Identification field for PROCNME runnme-link Identification field for RUNNME timetpl-link Identification field for TIMETPL usernme-link Identification field for USERNME priority Measure field- priority escalations Measure field- escalations lock-waits Measure field- number of loc
Entity and Attribute Definitions SQLSUM (continued) subvol Program subvolume filename Program file name run-unit SQL statement run unit statement-index SQL statement index aggregated-counters Aggregated counters follow: calls Number of calls elapsed-busy-time Amount of elapsed busy time records-used Number of records used records-accessed Number of records accessed disc-reads Number of disk reads messages Number of messages message-bytes Number of message bytes sorts Number of sort
Entity and Attribute Definitions SQLTOTAL (continued) recompiles Number of recompiles elapsed-recompile-time Amount of elapsed recompile time lock-waits Number of lock-waits timeouts Number of timeouts escalations Number of escalations SVNET Attribute Identification Definition svnet-link Identification field for SVNET cpu-num CPU number ctrl-type Measure field - product number, such as 3128 requests Measure field - requests to the SAC total-io-bytes Measure field - bytes transferred to
Entity and Attribute Definitions SVNET (continued) write-cbytes Measure field - message system protocol control bytes server-qtime Measure field - time to issue replies server-qlen-max Measure field - maximum number of items on the queue retries Measure field - tries before successful sends acks Measure field - explicit acknowledgment messages sent x-defrd-busy-time Measure field - wait time before switching from Y fabric to X fabric y-defrd-busy-time Measure field - wait time before switching
Entity and Attribute Definitions TERMINAL (continued) cpu-num Measure field- CPU number input-bytes-f Measure field - 64-bit version of input-bytes output-bytes-f Measure field - 64-bit version of output-bytes TOC Attribute Identification Definition loadid Identification field for LOAD load-id Redefines LOADID prefix-id Identification field for load prefix interval-id Identification field for load interval toc-type Table of contents type section-num Section number ret-name Name of repo
Entity and Attribute Definitions UNLINES (continued) device-info Device information device-type Device type device-subtype Device subtype counters Types of counters, as follows: read-busy-time Read busy time write-busy-time Write busy time requests Number of requests input-bytes Number of input bytes output-bytes Number of output bytes input-data-bytes Number of input data bytes output-data-bytes Number of output data bytes reads Number of reads writes Number of writes retries Nu
Entity and Attribute Definitions VSIOTOT volume File volume total-driver-calls Total number of driver calls total-dop-cache-hits Total number of DISCOPEN cache hits total-block-splits Total number of block splits total-reqs-blocked Total number of requests blocked remote-driver-calls Number of remote driver calls, input plus output remote-dop-cache-hits Number of remote DISCOPEN cache hits remote-block-splits Number of remote blocks splits remore-reqs-blocked Number of remote requests bloc
Entity and Attribute Definitions Tandem Performance Data Collector (TPDC) Manual— 526357-001 A -66
B Cross-Reference of Enform Queries to Data Files This example shows the XREFRPTS file, which cross-references Enform queries to the data files they use. !----------------------------------------------------------------------------! ! ! ! ! ! XREFRPTS ! ! ========== ! ! ! ! ! ! This file cross-references compiled ENFORM queries to the data files ! ! they require. You may edit this file to "customize" the selection ! ! and sequence of charts and/or reports contained within a named ! ! .
Cross-Reference of Enform Queries to Data Files !----------------------------------------------------------------------------! ! For each such line (unless, of course, it begins with an ! ! exclamation mark), TBUILD adds a "?Execute R" line to ! ! the command file it is building if, and only if: ! ! ! ! a. The ENFORM-compiled query file R exists and ! ! has a file code of 888, and ! ! ! ! b.
Cross-Reference of Enform Queries to Data Files CREATE CHARTS ! Name ! ------COVER PCOUNTS RCBCHRT ACUCHRT RDBCHRT DRTCHRT ADUCHRT DIOCHRT DCPCHRT TMFVIEW EXPAND DATACOM ! STACHRT WKLOAD AXLCHRT SQLPROG SQLSTAT ! PGSQLWT PGOVIEW PGOPENS PGLIOWT PGDIOWT PGSPLWT LDFIOWT DFSIOS DFPIOS BLKSPLT REQBLKD DFOATBL DFOSTBL DPMEMRY DFIODST PRIDIST ! DCUCHRT ( 80) Pseudo-graphic charts of overall activity/balance Uses [( ) = link optional] Description -------------------------- ---------------------------------GLOB
Cross-Reference of Enform Queries to Data Files CREATE REPORTS ! Name ! ------COVER PTIMLIN NEWPROC PATHWAY SERVICE ! SVCBYIO ! SERVLIO ! SERVPIO CPUREPT DPREPT DSCREPT PDISCIO CACHE1 TMFREPT SYSREPT NETREPT ! PALPHA ! PBSYOVR ! PCPUPRI PBSYCPU PMEMCPU PFLTCPU IOWOVER LIOBVOL ! LIOOVER SIOBVOL ! SIOOVER PIOBVOL ! PIOOVER SQLDETL ! HAYSTCK ! CACHE2 ! FILOPNS ! PROGLIO (132) Detailed reports on measured entities Uses [( ) = link optional] Description -------------------------- ----------------------------
C Enform Query Source File (QPGROUPS) This example shows the QPGROUPS file, which you can use in customizing some of your Enform source queries.
Enform Query Source File (QPGROUPS) =========================== Default Category Criteria ================ !--------------------------------------------------------------------! ! ! ! QPGROUPS ! ! ======== ! ! ! ! ?Sources: QSERVDEF ! ! ! !--------------------------------------------------------------------! ! ! ! This (partial) !ENFORM query source file is intended to be used ! ! to aid process categorization from MEASURE "PROCESS" data.
Enform Query Source File (QPGROUPS) ?Section CNT1 (count (1 over all where ( ?Section CNT2 ?Section CPU1 ?Section CPU2 ) ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section (sum ) ) (CPU-BUSY-TIME over all where ( ) ) ) DESCR01 ,PTYPE[ 1].PDESC := ("DISC Processes ") GROUP01 PRIORITY >= 219 and PROCESS-NAME <> "$IPB " DESCR02 ,PTYPE[ 2].
Enform Query Source File (QPGROUPS) ?Section GROUP07 ! ! ! ! ! ! ! ! ! ?Section ?Section ?Section ?Section ?Section ?Section or or or or or or or or or or or or or or or or or or or or or or or or FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME DESCR08 ,PTYPE[ 8].
Enform Query Source File (QPGROUPS) ! ! ! ! ! ! ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section ?Section or or or or or or FILENAME FILENAME FILENAME FILENAME FILENAME FILENAME DESCR11 ,PTYPE[11].PDESC GROUP11 FILENAME DESCR12 ,PTYPE[12].PDESC GROUP12 FILENAME DESCR13 ,PTYPE[13].PDESC GROUP13 FILENAME DESCR14 ,PTYPE[14].PDESC GROUP14 FILENAME ! or FILENAME ! or FILENAME DESCR15 ,PTYPE[15].PDESC GROUP15 ?Section ?Source QSERVDEF ?Section DESCR16 ,PTYPE[16].
Enform Query Source File (QPGROUPS) Tandem Performance Data Collector (TPDC) Manual— 526357-001 C- 6
Index A G Analysis 1-7 AUTO option 3-14 GPA 3-5 I B BEGIN option 3-15 C Charts, listed 3-25, 5-2, 5-5 COLLECT option 3-13 Components measured 1-6 TPDC 1-2, 3-1 Configuration hardware 1-6 software 1-6 Installation SPAM 2-2 TPDC 2-1 IOFACTOR option 3-17 L LINK attributes 1-5 M Data collection 1-5 DEFER option 2-7 Disk file labels 3-2 DWS file 1-4 Macros TBUILD 3-14 TGATHER 3-13 THELP 3-19 TPRODUCE 3-14 TSHOW 3-14 Measure entities 3-5 Measurement information 1-6 MINMSGS option 3-17 E N ENFORM com
Index O O R Options AUTO 3-14 BEGIN 3-15 COLLECT 3-13 DEFER 2-7 ENFORM 3-19 EXCLUDE 3-16 FILE 3-16 FLAG 3-16 IOFACTOR 3-17 MINMSGS 3-17 NAME 2-5, 3-3, 3-5, 3-9 NEEDLE 3-17 NOWAIT 2-5, 3-3, 3-6, 3-10 OUT 2-5, 3-3, 3-6, 3-10 PROGRAM 3-17 PROJECT 3-18 PURGE 3-18 REPORTS 2-6, 2-8, 3-3, 3-6, 3-10, 3-13 RESTART 2-8 RETAIN 2-5, 2-8 SHOW 3-14 SHOWRCV 3-18 STATS 3-18 SWAPVOL 3-13 VALIDATE 3-4 OUT option 2-5, 3-3, 3-6, 3-10 Output files TPDC 2-7 TPSFL 3-4 Reports listed 3-26, 4-1, 5-2, 5-3, 5-49 TPDC 1-7 REPORTS
