Change PEX Definition (CHGPEXDFN)
Where allowed to run: All environments (*ALL) Threadsafe: Conditional |
Parameters Examples Error messages |
The Change Performance Explorer Definition (CHGPEXDFN) command changes an existing Performance Explorer definition. The current values stored in an existing definition are inserted when the CHGPEXDFN command is prompted. Each definition is stored as a member in the QAPEXDFN file in library QUSRSYS. A Performance Explorer definition identifies the performance data that is to be collected during a Performance Explorer session. A session can be started using the STRPEX (Start Performance Explorer) command. When starting a new session, a Performance Explorer definition name must be provided.
Additional information about the Performance Explorer tool can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Restrictions:
- This command is shipped with public *EXCLUDE authority.
- You must have *EXECUTE authority to the library of each program specified on the PGM parameter.
- To use this command you must have *SERVICE special authority, or be authorized to the Service Trace function of the operating system. The Change Function Usage (CHGFCNUSG) command, with a function ID of QIBM_SERVICE_TRACE, can be used to change the list of users that are allowed to perform trace operations.
- The following user profiles have private authorities to use the command:
- QPGMR
- QSRV
- Two threads within the same job will not be allowed to run CHGPEXDFN at the same time. The thread that issued CHGPEXDFN first will run the command to completion while the second CHGPEXDFN waits.
Top |
Parameters
Keyword | Description | Choices | Notes |
---|---|---|---|
DFN | Definition | Name | Required, Key, Positional 1 |
LEVEL | Level | *SAME, *STANDARD, *SERVICE | Optional |
TYPE | Type | *SAME, *STATS, *TRACE, *PROFILE | Optional |
PRFTYPE | Profile type | *SAME, *PGM, *JOB, *SYSTEM | Optional |
JOB | Jobs | Single values: *SAME, *, *NONE Other values (up to 8 repetitions): Element list |
Optional |
Element 1: Job name | Qualified job name | ||
Qualifier 1: Job name | Generic name, name, *ALL | ||
Qualifier 2: User | Generic name, name, *ALL | ||
Qualifier 3: Number | 000001-999999, *ALL | ||
Element 2: Thread identifier |
Single values: *ALL, *SELECT Other values (up to 20 repetitions): 00000001-FFFFFFFF, *INITIAL |
||
Element 3: Subsystem | Generic name, name, *ALLSBS | ||
TASK | Task name | Single values: *ALL, *SAME, *NONE Other values (up to 8 repetitions): Generic name, name |
Optional |
PGM | Program to monitor | Values (up to 16 repetitions): Element list | Optional |
Element 1: Program | Qualified object name | ||
Qualifier 1: Program | Name, *ALL | ||
Qualifier 2: Library | Name, *SAME, *LIBL | ||
Element 2: Module | Name, *SAME, *ALL | ||
Element 3: Procedure | Character value, *SAME, *ALL | ||
Element 4: Type | *SAME, *PGM, *SRVPGM | ||
Element 5: Pane size | *SAME, 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, 4096 | ||
DTAORG | Data organization | *FLAT, *HIER, *SAME | Optional |
MAXSTG | Maximum storage to use | Integer, *SAME | Optional |
TRCFULL | Trace full | *SAME, *STOPTRC, *WRAP | Optional |
MRGJOB | Merge job data | *SAME, *YES, *NO | Optional |
ADDTHDOPT | Add threads/tasks option | *SAME, *ALL, *NEW, *CURRENT | Optional |
LSTALLJOB | List all jobs/tasks | *SAME, *NO, *YES | Optional |
PMC | PMC selection | Element list | Optional |
Element 1: Group number | 1-9999, *SAME | ||
Element 2: PMC number | 1-6, *SAME | ||
Element 3: Scope | *SAME, *TASK, *PARTITION, *SYSTEM | ||
Element 4: MMCR0 value (hex) | 0000000000000000-FFFFFFFFFFFFFFFD, *SAME, *NONE | ||
Element 5: MMCR1 value (hex) | 0000000000000000-FFFFFFFFFFFFFFFD, *SAME, *NONE | ||
Element 6: MMCR2 value (hex) | 0000000000000000-FFFFFFFFFFFFFFFD, *SAME, *NONE | ||
Element 7: MMCRA value (hex) | 0000000000000000-FFFFFFFFFFFFFFFD, *SAME, *NONE | ||
INTERVAL | Sampling interval | Element list | Optional |
Element 1: Milliseconds | 0.1-10000.0, *SAME, *NONE | ||
Element 2: Randomize | *SAME, *FIXED, *VARY | ||
ITVSPEC | Sampling interval | *SAME, *TIME, *COUNT | Optional |
ITVTIME | Sampling interval time | Element list | Optional |
Element 1: Milliseconds | 0.1-2000000000.0, *SAME, *NONE | ||
Element 2: Randomize | *SAME, *FIXED, *VARY | ||
ITVCOUNT | Sampling interval count | Element list | Optional |
Element 1: Count | 1-2000000000, *SAME, *NONE | ||
Element 2: Randomize | *SAME, *FIXED, *VARY | ||
PGMBKTEVT | Program bracketing events | Single values: *DFT Other values (up to 3 repetitions): *SAME, *MISTREND, *MIENTRYEXIT, *JVA, *PRC |
Optional |
TRCTYPE | Trace type | Single values: *SLTEVT Other values (up to 12 repetitions): *SAME, *CALLRTN, *BASIC, *DSKIO1, *DSKIO2, *DSKSVR, *DSKSTG, *VRTADR, *PGMACT, *FILEOPEN, *PRFDTA, *TASKSWT, *HEAP |
Optional |
SLTEVT | Specific events | *SAME, *NO, *YES | Optional |
MCHINST | Machine instructions | Single values: *SAME, *ALL, *NONE Other values (up to 50 repetitions): Character value |
Optional |
PGMEVT | Program events | Single values: *SAME, *NONE Other values (up to 10 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *MIENTRY, *MIEXIT, *MISTR, *MIEND, *JVAENTRY, *JVAEXIT, *PRCENTRY, *PRCEXIT, *PASEPRCENTRY, *PASEPRCEXIT | ||
Element 2: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
BASEVT | Base events | Single values: *SAME, *NONE Other values (up to 34 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *PRCCRT, *PRCDLT, *TASKCRT, *TASKDLT, *TASKSWTIN, *TASKSWTOUT, *TASKSWTOUTQ, *TASKSWTOUTINT, *TASKAVAIL, *TASKSMTCHANGE, *PMCO, *PGMDBG, *ACTGRPACTPGM, *ACTGRPCRT, *ACTGRPDLT, *EXCP, *MIEXCP, *DCRINTSTR, *DCRINTEND, *CPUVRYON, *CPUVRYOFF, *CPUSWTIN, *CPUSWTOUTY, *CPUSWTOUTP, *CPUAVAIL, *CPUSWT, *SETPRF, *WKLDCAPPING, *PRCFLDSUSPEND, *PRCFLDRESUME, *LPARSUSPEND, *LPARRESUME, *ACTDTA, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
STGEVT | Storage events | Single values: *SAME, *NONE Other values (up to 13 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *CRTSEG, *DLTSEG, *EXDSEG, *FNDSEGSIZ, *TRUNCSEG, *SYSHEAP, *RESHEAP, *LCLHEAP, *USRHEAP, *ACTGRPHEAP, *HDLHEAP, *CLEHEAP, *STGPTC, *TSADRFALSETRAP, *SPCADRFALSETRAP, *CHGSEGATR, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
DSKEVT | Disk events | Single values: *SAME, *NONE Other values (up to 15 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *ALLSTR, *READSTR, *READEND, *WRTSTR, *WRTEND, *PGREADSTR, *PGREADEND, *PGWRTSTR, *PGWRTEND, *RMTWRTSTR, *RMTWRTEND, *RMTPGWRTSTR, *RMTPGWRTEND, *LOGSNSEND, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
FAULTEVT | Fault events | Single values: *SAME, *NONE Other values (up to 4 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *STR, *ENDOK, *ENDERR, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
JOBEVT | Job events | Single values: *SAME, *NONE Other values (up to 10 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *LWSTR, *LWEND, *INELIGIBLE, *ACTIVE, *INTERRUPT, *MPLPOOLCHG, *TOBCHMPLPOOL, *TSLEND, *MPLLEAVE, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
LCKEVT | Lock events | Single values: *SAME, *NONE Other values (up to 5 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *LWEND, *OBJLOCK, *SPCLOCK, *DBLOCK, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
SAREVT | SAR events | Single values: *SAME, *NONE Other values (up to 18 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *ALLSTR, *CLR, *CLRPIN, *EXCHCLR, *READ, *READASYNC, *READASYNCCALL, *READPIN, *EXCHREAD, *EXCHREADASYNC, *WRT, *WRTASYNC, *WRTRMV, *WRTPGOUT, *RMV, *UNPINRMV, *ENDOK, *ENDERR, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
EXPCCHEVT | Expert cache events | Single values: *SAME, *NONE Other values (up to 3 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *TUNE, *TUNEDB, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
DSKSVREVT | Disk server events | Single values: *SAME, *NONE Other values (up to 11 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *CTLEND, *RQSACP, *RQSRJT, *SVRTASKSTR, *SVRTASKEND, *RQSRCV, *RQSENDOK, *RQSENDERR, *TAPSNDCMD, *TAPDONE, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
OSEVT | Operating System events | Single values: *SAME, *NONE Other values (up to 54 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *DBIO, *DBOPEN, *DBSVRCNN, *DBSVRREQ, *DIRSRV, *DTAARA, *DTAQ, *HOSTSVRCNN, *IFSCOMP, *IFSIO, *IFSOPEN, *MIEV8, *MGTC, *TRCCOMP, *TRCDTA, *USRTNS, *SAVRST, *ADDTHD1, *ARMTRC, *MIEV25, *MIEV26, *MIEV27, *MIEV28, *MIEV29, *MIEV30, *MIEV35, *MIEV36, *MIEV37, *MIEV38, *MIEV39, *MIEV40, *MIEV41, *MIEV42, *MIEV43, *MIEV44, *MIEV45, *MIEV46, *MIEV47, *MIEV48, *MIEV49, *MIEV50, *MIEV51, *MIEV52, *MIEV53, *MIEV54, *MIEV55, *MIEV56, *MIEV57, *MIEV58, *MIEV59, *MIEV60, *MIEV61, *MIEV62, *MIEV63, *MIEV64, *MIEV65, *MIEV66, *MIEV67, *MIEV68, *MIEV69, *MIEV70, *MIEV71, *MIEV72, *MIEV73, *MIEV74, *MIEV75, *MIEV76, *MIEV77, *MIEV78, *MIEV79, *MIEV80, *MIEV81, *MIEV82, *MIEV83, *MIEV84, *MIEV85, *MIEV86, *MIEV87, *MIEV88, *MIEV89, *MIEV90, *MIEV91, *MIEV92, *MIEV93 | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
JVAEVT | Java events | Single values: *SAME, *NONE Other values (up to 20 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *OBJCRT, *GBGCOL, *GBGCOLSWEEP, *THDCRT, *THDDLT, *THDWAIT, *THDNFY, *THDNFYALL, *CLSLOAD, *CLSUNLOAD, *LIBOPR, *TFMSTR, *LCKSTR, *UNLCK, *JVAEXCP, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
CMNEVT | Communications events | Single values: *SAME, *NONE Other values (up to 6 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *SOCKETSAPI, *SOCKETS, *IP, *TCP, *TCPIN, *TCPOUT, *UDP, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
APPEVT | Application events | Single values: *SAME, *NONE Other values (up to 7 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *DOMTRCDTA, *WAS, *CONNECT, *APPEVT1, *APPEVT2, *APPEVT3, *APPEVT4 | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
PASEEVT | PASE events | Single values: *SAME, *NONE Other values (up to 22 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *SYSCALLSTR, *SYSCALLEND, *FORKPARENTSTR, *FORKCHILDSTR, *FORKEND, *PRCINITSTR, *PRCINITEND, *EXECSTR, *EXECEND, *EXIT, *THDINITSTR, *THDINITEND, *THDEND, *LOADSTR, *LOADEND, *UNLOADSTR, *UNLOADEND, *SIGPRCSND, *SIGTHDSND, *SIGRCV, *TRCHOOK, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
FILSVREVT | File server events | Single values: *SAME, *NONE Other values (up to 9 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *IFSBSF, *IFSDIR, *IFSLFS, *LIC, *NETSVR, *NFS, *OS, *VNODMGMT, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
SYNCEVT | Synchronization events | Single values: *SAME, *NONE Other values (up to 13 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *CONDSET, *CONDWAIT, *HDLMTXRLS, *HDLMTXWAIT, *PTRMTXLOCK, *PTRMTXUNLOCK, *PTRSEMPOST, *PTRSEMWAIT, *NAMSEMPOST, *NAMSEMWAIT, *TKNLOCK, *TKNUNLOCK, *MTXCLEANUP, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
JRNEVT | Journal events | Single values: *SAME, *NONE Other values (up to 10 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *STRCOMMIT, *ENDCOMMIT, *STROBJFORCE, *ENDOBJFORCE, *STRROLLBACK, *ENDROLLBACK, *STRCYCLE, *STREVAL, *ENDEVAL, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
DBEVT | Database events | Single values: *SAME, *NONE Other values (up to 4 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *DASDMETER, *IO, *OPTIMIZER, *OPTIMIZER2, *RUNTIME, *SEGMENTCACHE, *SEIZELOCK, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
RSCAFNEVT | Resources affinity events | Single values: *SAME, *NONE Other values (up to 9 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *CHGTASK, *CHGGRP, *CHGCFG, *BALCFG, *CHKBAL, *VFYBAL, *ANZBAL, *CHGBALSTT, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
VRTIOEVT | Virtual I/O events | Single values: *SAME, *NONE Other values (up to 12 repetitions): Element list |
Optional |
Element 1: Event identifier | *ALL, *SCSICLIENT, *SCSISERVER, *ISCSI, *SCSICMD, *SCSITSKMGMT, *DISKSTR, *DISKEND, *OPTSTR, *OPTEND, *TAPSTR, *TAPEND, *ETHADPT, *SERVICE | ||
Element 2: Event counter | *NONE, 1, 2, 3, 4, 5, 6, 7, 8 | ||
Element 3: Event format | *FORMAT1, *FORMAT2, *FORMAT3, *FORMAT4 | ||
TEXT | Text 'description' | Character value, *SAME, *BLANK | Optional |
Top |
Definition (DFN)
Specifies the name of the Performance Explorer definition to be changed.
This is a required parameter.
- name
- Specify the name of the Performance Explorer definition to change.
Top |
Level (LEVEL)
Specifies the level of the PEX definition.
- *SAME
- The value does not change.
- *STANDARD
- This creates standard PEX definitions. Use this unless you have been directed by the IBM Service organization to use LEVEL(*SERVICE).
- *SERVICE
- This creates advanced PEX definitions. It contains several parameters that are not available when using LEVEL(*STANDARD). Use this ONLY if you have been told to do so by the IBM Service organization. Misuse of the Service level could create PEX definitions that could severely degrade the performance of your system, when used by the STRPEX command.
Top |
Type (TYPE)
Specifies the type of performance data to be collected.
- *SAME
- The value does not change.
- *STATS
- General performance program statistics are collected to help identify problem areas. This mode is mainly used as a map to help determine if and where more detailed information should be collected and analyzed.
In addition to collecting the information specified above, TYPE(*STATS) also provides the option of counting the occurrences of specific types of events. These can be counted in any of 4 counters provided.
To count the occurrences of a particular type of event, specify SLTEVT(*YES) and then choose the event-identifier(s) to be counted from the various event categories.
For example, TYPE(*STATS) SLTEVT(*YES) JOBEVT((*ALL 1)) would count all job events in counter 1.
- *TRACE
- Detailed trace information is collected. This is the most detailed type of performance data collection available.
- *PROFILE
- Specific programs are sampled to identify sections of code that are using larger amounts of resources. This information is very valuable when the user wants to improve the performance of a specific program or application.
Top |
Profile type (PRFTYPE)
Specifies the type of profile to collect. This parameter is only valid if TYPE(*PROFILE) is specified.
- *PGM
- Specific programs are sampled to identify sections of code that are using larger amounts of resources. This information is very valuable when the user wants to improve the performance of a specific program or application.
- *JOB
- Specific jobs are sampled to identify programs and procedures that are using larger amounts of resources. This mode can provide a view of all the programs and procedures in the entire system, and is equivalent to *TRACE mode of *PMCO events with a specified interval.
Top |
Job name (JOB)
Specifies which jobs are included in the Performance Explorer data collection session.
Single values
- *SAME
- The value does not change.
- *
- Only the job that issues the STRPEX (Start Performance Explorer) command is included.
Element 1: Job name
Qualifier 1: Job name
- *ALL
- All jobs on the system are included.
- name
- Specify the name of the job to be included in the Performance Explorer data collection session.
- generic-name
- Specify the generic name of the job to be included. A generic name is a character string of one or more characters followed by an asterisk (*). A generic name specifies all items that have names with the same prefix as the generic name.
Qualifier 2: User
- *ALL
- All jobs that match the specified job name are included.
- name
- Specify the name of the user of the job to be included.
- generic-name
- Specify the generic user name of the jobs to be included.
Qualifier 3: Number
- *ALL
- All jobs that match the specified job name and user name are included.
- job-number
- Specify the job number to further qualify the job name and user name.
Element 2: Thread identifier
- *ALL
- All threads of the specified job are included.
- *INITIAL
- Only the initial thread of the specified job is included.
- *SELECT
- Select the threads from a list of threads for the specified job. This value is only valid if the command is run in an interactive job.
- thread-identifier
- Specify the thread identifier of the job to be included. This is the thread ID as shown by the WRKJOB command.
Element 3: Subsystem
- *ALLSBS
- All subsystems are searched to find jobs matching the specified job name, user, and number.
- name
- Specify the subsystem name of the jobs to be included.
- generic-name
- Specify the generic subsystem name of the jobs to be included.
Top |
Task name (TASK)
Specifies which licensed internal code (LIC) tasks are included in the Performance Explorer data collection session. Up to 8 task names may be specified.
Note: LIC tasks can be obtained from the Performance Tools reports and WRKSYSACT command. There is no guarantee that LIC task names will remain the same or exist from system to system or release to release.
Single values
- *SAME
- The value does not change.
- *NONE
- No LIC tasks on the system are included.
- *ALL
- All LIC tasks on the system are included.
Other values (up to 8 repetitions)
- name
- Specify the name of the task to be included in the Performance Explorer data collection session.
- generic-name
- Specify the generic name of the task to be included. A generic name is a character string of one or more characters followed by an asterisk (*). A generic name specifies all items that have names with the same prefix as the generic name.
Top |
Program to monitor (PGM)
Specifies the program whose performance profile data is collected. This parameter is only valid if TYPE(*PROFILE) is specified. Up to 16 programs may be specified.
Element 1: Program
Qualifier 1: Program
- *SAME
- The value does not change.
- *ALL
- All programs in the specified library are sampled.
- name
- Specify the name of the program to be sampled.
Qualifier 2: Library
- *SAME
- The value does not change.
- *LIBL
- The library list of the job that issues the STRPEX command is searched to find the specified program or service program.
- name
- Specify the library which contains the program or service program.
Element 2: Module
- *SAME
- The value does not change.
- *ALL
- All modules in the program or service program will be sampled. If sampling an OPM (Original Program Model) program, specify *ALL for this element.
- name
- Specify the name of the module within the program or service program that is to be sampled.
Element 3: Procedure
- *SAME
- The value does not change.
- *ALL
- All procedures in the specified module are sampled.
- name
- Specify the name of a procedure within the specified module that is to be sampled. Specify the procedure name within single quotes if the procedure name contains lower case characters.
Element 4: Type
- *SAME
- The value does not change.
- *PGM
- The program being specified is a program (*PGM) object.
- *SRVPGM
- The program being specified is a service program (*SRVPGM) object.
Element 5: Pane size
The pane size is the number of consecutive program instruction addresses assigned to each counter. The smaller the pane size, the more fine-grained the program profile information will be.
- *SAME
- The value does not change.
- number
- Specify the pane size to use for the program. Valid values are 4, 8, 16, 32, 64, 128, 256, 512, 1024, 2048, and 4096.
Top |
Data organization (DTAORG)
Specifies how the data is organized.
Note: This parameter is only valid if TYPE(*STATS) is specified.
- *SAME
- The value does not change.
- *FLAT
- The Performance Explorer tool will not collect data for a parent-child relationship.
- *HIER
- The Performance Explorer tool will collect data for a parent-child relationship.
Top |
Maximum storage to use (MAXSTG)
Specifies the maximum amount of storage, in kilobytes (K), that the Performance Explorer tool uses for collecting trace data.
Note: This parameter is only valid if TYPE(*TRACE) is specified.
- *SAME
- The value does not change.
- maximum-K-bytes
- Specifies the requested maximum amount of storage, in kilobytes (K), to use for the collected trace records. The system calculates the minimum amount of storage that is necessary for the trace. This minimum storage size calculation depends on the system's processor configuration. The minimum amount of storage may be significantly larger than the size specified on the MAXSTG parameter. The system uses the larger of the two values.
Top |
Trace full (TRCFULL)
Specifies whether the trace records wrap (replace oldest records with new records) or whether the trace stops when all of the storage specified by the MAXSTG parameter has been used.
Note: This parameter is only valid if TYPE(*TRACE) is specified.
- *SAME
- The value does not change.
- *STOPTRC
- Tracing stops when the trace file is full of trace records.
- *WRAP
- When the trace file is full, the trace wraps to the beginning. The oldest trace records are written over by new ones as they are collected.
Top |
Merge job data (MRGJOB)
Specifies whether the data from different jobs should be merged together or kept separate.
Note: This parameter is only valid if DTAORG(*FLAT) is specified.
- *SAME
- The value does not change.
- *YES
- The data from individual jobs is merged.
- *NO
- The data from individual jobs is kept separate.
Top |
Add threads/tasks option (ADDTHDOPT)
Specifies what types of threads and tasks should be included in the Performance Explorer session based on the creation time of the threads and tasks relative to the start time of the Performance Explorer session.
- *SAME
- The value does not change.
- *ALL
- The collection will include all threads and tasks that satisfy the criteria specified by the JOB and TASK parameters.
- *NEW
- The collection will include only new threads and tasks that satisfy the criteria specified by the JOB and TASK parameters and are created after the Performance Explorer session is started. Any threads or tasks that are active at the time the Performance Explorer session is started will not be included.
- *CURRENT
- The collection will include only threads and tasks that satisfy the criteria specified by the JOB and TASK parameters and are active at the time the Performance Explorer session is started. Any threads or tasks created after the start of the Performance Explorer session will not be included.
Top |
List all jobs/tasks (LSTALLJOB)
Specifies whether summary data should be listed for all jobs and tasks in the system, or only those jobs and tasks specified on the JOB or TASK parameter.
- *SAME
- The value does not change.
- *NO
- Data for a job or task is listed in the task information file only if that job or task is specified on the JOB or TASK parameter.
- *YES
- Data for all jobs and tasks is listed in the task information file, even if the job or task is not specified on the JOB or TASK parameter.
Top |
PMC selection (PMC)
Specifes the hardware events that will be counted in the PMC registers and the PMC register that will be used for sampling.
Note: This parameter is valid only when LEVEL(*SERVICE) is specified.
Element 1: Group number
- *SAME
- The value does not change.
- 1-9999
- Selects the PMC group number within the PMC table. The valid values are dependent on the hardware platform.
Element 2: PMC number
Selects the PMC register to be used for sampling. This value, combined with the group number, will define a specific hardware event condition to be used for sampling.
- *SAME
- The value does not change.
- 1-6
- Specifies the PMC register by number. The valid values are 1-6.
Element 3: Scope
- *SAME
- The value does not change.
- *TASK
- The sampling intervals are managed separately in each task using a virtualized representation of the PMC registers. When the sampling interval is reached within a task, a sample is taken.
- *PARTITION
- The sampling intervals are managed in each hardware thread within the partition. When the sampling interval is reached, a sample is taken. The resulting samples will show more of the internal system contributions to events being sampled than can be seen with the *TASK option.
Element 4: MMCR0 value (hex)
- *SAME
- The value does not change.
- *NONE
- Do not override the value of the MMCR0 register. Use the value defined by the specified Group number.
- 0000000000000000-FFFFFFFFFFFFFFFE
- Hex value to be loaded into the MMCR0 register. This value will override the value defined by the specified group number.
Element 5: MMCR1 value (hex)
- *SAME
- The value does not change.
- *NONE
- Do not override the value of the MMCR1 register. Use the value defined by the specified Group number.
- 0000000000000000-FFFFFFFFFFFFFFFE
- Hex value to be loaded into the MMCR1 register. This value will override the value defined by the specified group number.
Element 6: MMCR2 value (hex)
- *SAME
- The value does not change.
- *NONE
- Do not override the value of the MMCR2 register. Use the value defined by the specified Group number.
- 0000000000000000-FFFFFFFFFFFFFFFE
- Hex value to be loaded into the MMCR2 register. This value will override the value defined by the specified group number.
Element 7: MMCRA value (hex)
- *SAME
- The value does not change.
- *NONE
- Do not override the value of the MMCRA register. Use the value defined by the specified Group number.
- 0000000000000000-FFFFFFFFFFFFFFFE
- Hex value to be loaded into the MMCRA register. This value will override the value defined by the specified group number.
Top |
Sampling interval (INTERVAL)
Specifies the interval at which samples are taken of the program. A small interval will cause a high number of samples to be taken, and will also cause higher overhead. The smaller the sampling interval, the larger the amount of trace data collected.
Note: This parameter is valid only when LEVEL(*STANDARD) and either TYPE(*PROFILE) or TYPE(*TRACE) are specified.
Element 1: Milliseconds
- *SAME
- The value does not change.
- *NONE
- Interval is not used.
- 0.1-10,000.0
- Specify the interval at which samples of the program are taken. Valid values range from 0.1 to 10,000.0 milliseconds.
Element 2: Randomize
- *SAME
- The value does not change.
- *FIXED
- The sampling interval will always be the value specified for element 1 of this parameter.
- *VARY
- The sampling interval will be approximately the value specified for element 1 of this parameter. The specified interval will be changed each time by adding or subtracting a small random percentage of the interval. This may be necessary to eliminate harmonics when the set of threads/tasks in a collection have settled into a steady state sequence of machine cycles that do not vary in order or length during the collection.
Top |
Sampling interval units (ITVSPEC)
Specifies the units to be used for the sampling interval.
- *TIME
- A time in milliseconds is used to specify the sampling interval. The time is specified with the ITVTIME keyword.
- *COUNT
- A count is used to specify the sampling interval. The count is specified with the ITVCOUNT keyword.
Note: This parameter is valid only when LEVEL(*SERVICE) and either TYPE(*PROFILE) or TYPE(*TRACE) are specified.
Top |
Sampling interval time (ITVTIME)
Specifies the interval at which samples are taken of the program. A small interval will cause a high number of samples to be taken, and will also cause higher overhead. The smaller the sampling interval, the larger the amount of trace data collected.
Note: This parameter is valid only when LEVEL(*SERVICE) and either TYPE(*PROFILE) or TYPE(*TRACE) are specified.
Element 1: Milliseconds
- *SAME
- The value does not change.
- *NONE
- Interval is not used.
- 0.1 - 10,000.0
- Specify the time interval (in milliseconds) at which samples of the program are taken. Valid values range from 0.1 - 10,000.0.
Element 2: Randomize
- *SAME
- The value does not change.
- *FIXED
- The sampling interval will always be the value specified for element 1 of this parameter.
- *VARY
- The sampling interval will be approximately the value specified for element 1 of this parameter. The specified interval will be changed each time by adding or subtracting a small random percentage of the interval. This may be necessary to eliminate harmonics when the set of threads/tasks in a collection have settled into a steady state sequence of machine cycles that do not vary in order or length during the collection.
Top |
Sampling interval count (ITVCOUNT)
Specifies the interval at which samples are taken of the program. A small interval will cause a high number of samples to be taken, and will also cause higher overhead. The smaller the sampling interval, the larger the amount of trace data collected.
Note: This parameter is valid only when LEVEL(*SERVICE) and either TYPE(*PROFILE) or TYPE(*TRACE) are specified.
Element 1: Count
- *SAME
- The value does not change.
- *NONE
- Interval is not used.
- 1 - 2,147,483,648
- Specifies a count. Valid values range from 1 - 2,147,483,648.
Element 2: Randomize
- *SAME
- The value does not change.
- *FIXED
- The sampling interval will always be the value specified for element 1 of this parameter.
- *VARY
- The sampling interval will be approximately the value specified for element 1 of this parameter. The specified interval will be changed each time by adding or subtracting a small random percentage of the interval. This may be necessary to eliminate harmonics when the set of threads/tasks in a collection have settled into a steady state sequence of machine cycles that do not vary in order or length during the collection.
Top |
Program bracketing events (PGMBKTEVT)
Specifies which program call flow events are included in the TYPE(*STATS) definition.
Note: This parameter is only valid if TYPE(*STATS) is specified.
Single values
- *SAME
- The value does not change.
- *DFT
- Statistics are to be collected on *MISTREND events, *MIENTRYEXIT events for programs and service programs, and *JVA events.
Other values (up to 3 repetitions)
- *MISTREND
- Statistics are to be collected on all machine instructions.
- *MIENTRYEXIT
- Statistics are to be collected on programs and procedures. This includes any program that has been compiled at optimization level 30 or below. Optimization level 40 programs are also enabled, but only for procedures that stack a frame on the invocation stack when called (non-leaf procedures).
- *JVA
- Statistics are to be collected on Java methods. This includes interpreted Java and Java programs running in JIT mode where the property 'os400.enbpfrcol=1' has been set. This is only for classic VM, not IBM Technology for Java.
- *PRC
- *PRC will behave exactly like *MIENTRYEXIT.
Top |
Trace type (TRCTYPE)
Specifies what type of trace performance data to be collected. Additional information about the TRCTYPE options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if TYPE(*TRACE) is specified.
Single values
- *SAME
- The value does not change.
- *SLTEVT
- Only selected individual events (xxxEVT parameters) and machine instructions (MCHINST parameter) are included in the trace definition.
Note: If TRCTYPE(*SLTEVT) is specified, SLTEVT(*YES) must also be specified.
Other values (up to 12 repetitions)
- *CALLRTN
- Specifies that call return events are included in the trace definition. Call return events occur when a program is entered and exited as well as when certain machine instructions are started and completed.
- *BASIC
- Specifies that events relative to general performance analysis are included in the trace definition. This option should be used when it is unclear as to what type of performance problem determination is necessary.
- *DSKIO1
- Specifies that events associated with disk input/output operations are included in the trace definition.
- *DSKIO2
- Specifies that events associated with disk input/output operations plus higher level requests to do input/output operations are included in the trace definition.
- *DSKSVR
- Specifies that events associated with disk server operations are included in the trace definition.
- *DSKSTG
- Specifies that events associated with disk storage consumption are included in the trace definition.
- *VRTADR
- Specifies that events associated with virtual address assignment are included in the trace definition.
- *PGMACT
- Specifies that events associated with program activations and deactivations are included in the trace definition.
- *FILEOPEN
- Specifies that events associated with file (*FILE) opens are included in the trace definition.
- *PRFDTA
- Specifies that events associated CPU instruction profiling are included in the trace definition.
- *TASKSWT
- Specifies that events associated with tasking are included in the trace definition.
Top |
Specific events (SLTEVT)
For trace mode (TYPE(*TRACE)) collections, SLTEVT allows individual machine instructions and events to be specified in addition to the categories of events available with the TRCTYPE parameter. For statistics mode (TYPE(*STATS)) collections, SLTEVT allows individual events to be counted.
Note: This parameter is only valid if TYPE(*TRACE) or TYPE(*STATS) is specified.
- *SAME
- The value does not change.
- *NO
- Do not allow selection of specific events.
- *YES
- Allow selection of specific events.
Top |
Machine instructions (MCHINST)
Specifies which machine instructions are included in the Performance Explorer data collection session. Use this parameter in conjunction with the PGMEVT parameter. MCHINST controls which of the possible machine instructions are to be included in the collection. PGMEVT(*MISTR and *MIEND) controls whether any machine instructions are included in the collection.
Note: This parameter is only valid if TYPE(*TRACE) and SLTEVT(*YES) is specified.
- *SAME
- The value does not change.
- *ALL
- All machine instructions that are available for collection are included.
- *NONE
- No machine instructions available for collection are included.
Other values (up to 50 repetitions)
- name
- Specify the name of the machine instruction to be included in the Performance Explorer data collection session.
Note: Specifying machine instructions should be done only under the direction of your service representative.
Top |
Program events (PGMEVT)
Specifies which program call flow events are included in the trace definition. Additional information about the PGMEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if TYPE(*TRACE) and SLTEVT(*YES) are specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No program call flow events are included in the trace definition.
- *ALL
- All program call flow events are included in the trace definition.
Element 1: Event identifier
- event-identifier
- Specify the call flow event identifier to be included in the trace definition.
Note: The *MIENTRY, *MIEXIT, *PRCENTRY and *PRCEXIT events are always available in programs that have been compiled at optimization level 30 or below. For optimization level 40, these events are available only for the procedures that stack a frame on the invocation stack when called (non-leaf procedures).
The *JVAENTRY and *JVAEXIT are available in interpreted Java and Java programs running in JIT mode where the property 'os400.enbpfrcol=1' has been set. This is only for classic VM, not IBM Technology for Java.
Element 2: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Base events (BASEVT)
Specifies which base events are included in the definition. Additional information about the BASEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No base events are included in the definition.
- *ALL
- All base events are included in the trace mode definition, or counted in the statistics mode definition. If TYPE(*STATS), you can specify which event-identifiers are to be counted. Some of the base event event-identifiers cannot be counted. An informational message will be shown if a definition attempts to count all base events.
Element 1: Event identifier
- event-identifier
- Specify the base event identifier to be traced or counted.
Element 2: Event counter
- *NONE
- No base events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events. Values other than *FORMAT1 are valid for only the *PMCO and *SWOQ events. For all other events, *FORMAT1 will be used regardless of what format is specified.
*FORMAT2 for the *PMCO event will collect the same data as *FORMAT1 plus an additional 16 levels of the call stack.
*FORMAT3 for the *PMCO event will collect the same data as *FORMAT1 plus 4 additional PMC registers.
*FORMAT2 for the *SWOQ event will collect the same data as *FORMAT1 plus an additional 16 levels of the call stack.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Storage events (STGEVT)
Specifies which auxiliary storage management events are included in the definition. Additional information about the STGEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No auxiliary storage management events are included in the definition.
- *ALL
- All auxiliary storage management events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the auxiliary storage management event identifier to be included.
Element 2: Event counter
- *NONE
- No auxiliary storage management events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Disk events (DSKEVT)
Specifies which disk events are included in the definition. Additional information about the DSKEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No disk events are included in the definition.
- *ALL
- All disk events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- *ALLSTR
- All disk start events are traced, if TYPE(*TRACE), or counted, if TYPE(*STATS).
- event-identifier
- Specify the disk event identifier to be included.
Element 2: Event counter
- *NONE
- No disk events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Fault events (FAULTEVT)
Specifies which page fault events are included in the definition. Additional information about the FAULTEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No page fault events are included in the definition.
- *ALL
- All page fault events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the page fault event identifier to be included.
Element 2: Event counter
- *NONE
- No page fault events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Job events (JOBEVT)
Specifies which job or process related events are included in the definition. Additional information about the JOBEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No job or process related events are included in the definition.
- *ALL
- All job or process related events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the job or process related event identifier to be included.
Element 2: Event counter
- *NONE
- No job or process related events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Lock events (LCKEVT)
Specifies which lock or seize events are included in the definition. Additional information about the LCKEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No lock or seize events are included in the definition.
- *ALL
- All lock or seize events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the lock or seize event identifier to be included.
Element 2: Event counter
- *NONE
- No lock or seize events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
SAR events (SAREVT)
Specifies which segment address register events are included in the definition. Additional information about the SAREVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No segment address register events are included in the definition.
- *ALL
- All segment address register events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the segment address register event identifier to be included.
Element 2: Event counter
- *NONE
- No segment address register events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Expert cache events (EXPCCHEVT)
Specifies which expert cache events are included in the definition. Additional information about the EXPCCHEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No expert cache events are included in the definition.
- *ALL
- All expert cache events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the expert cache event identifier to be included.
Element 2: Event counter
- *NONE
- No expert cache events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Disk server events (DSKSVREVT)
Specifies which disk server events are included in the definition. Additional information about the DSKSVREVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No disk server events are included in the definition.
- *ALL
- All disk server events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the disk server event identifier to be included.
Element 2: Event counter
- *NONE
- No disk server events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Operating System events (OSEVT)
Specifies which operating system events are included in the definition. Additional information about the OSEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No operating system events are included in the definition.
- *ALL
- All operating system events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the operating system event identifier to be included.
Element 2: Event counter
- *NONE
- No operating system events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Java events (JVAEVT)
Specifies which Java events are included in the definition. These events are only for classic VM, not IBM Technology for Java. Additional information about the JVAEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No Java events are included in the definition.
- *ALL
- All Java events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the Java event identifier to be included.
Element 2: Event counter
- *NONE
- No Java events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
*FORMAT2 for the *OBJCRT, *LCKSTR, and *UNLCK events will collect the same data as *FORMAT1 plus an additional 16 levels of the call stack.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Communications events (CMNEVT)
Specifies which communications events are included in the definition. Additional information about the CMNEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No communications events are included in the definition.
- *ALL
- All communications events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the communications event identifier to be included.
Element 2: Event counter
- *NONE
- No communications events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events. Values other than *FORMAT1 are valid for only the *SOCKETS events. For all other events, *FORMAT1 will be used regardless of what format is specified.
*FORMAT2 for the *SOCKETS event will collect the same data as *FORMAT1 except the application data contains 128 bytes of data instead of 64.
*FORMAT3 for the *SOCKETS event will collect the same data as *FORMAT1 except the application data contains 256 bytes of data instead of 64.
*FORMAT4 for the *SOCKETS event will collect the same data as *FORMAT1 except the application data contains 500 bytes of data instead of 64.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Application events (APPEVT)
Specifies which application events are included in the definition. Additional information about the APPEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No application events are included in the definition.
- *ALL
- All application events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the application event identifier to be included.
Element 2: Event counter
- *NONE
- No application events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
PASE events (PASEEVT)
Specifies which Portable Application Solution Environment (PASE) events are included in the definition. Additional information about the PASEEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No PASE events are included in the definition.
- *ALL
- All PASE events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the PASE event identifier to be included.
Element 2: Event counter
- *NONE
- No PASE events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
File server events (FILSVREVT)
Specifies which NetServer, File Server and Network File System Server and Client events are included in the definition. Additional information about the FILSVREVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No NetServer, File Server and Network File System Server and Client events are included in the definition.
- *ALL
- All NetServer, File Server and Network File System Server and Client events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the NetServer, File Server and Network File System Server and Client event identifier to be included.
Element 2: Event counter
- *NONE
- No NetServer, File Server and Network File System Server and Client events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Synchronization events (SYNCEVT)
Specifies which synchronization events are included in the definition. Additional information about the SYNCEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No synchronization events are included in the definition.
- *ALL
- All synchronization events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the synchronization event identifier to be included.
Element 2: Event counter
- *NONE
- No synchronization events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Journal events (JRNEVT)
Specifies which journal events are included in the definition. Additional information about the JRNEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No journal events are included in the definition.
- *ALL
- All journal events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the journal event identifier to be included.
Element 2: Event counter
- *NONE
- No journal events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Database events (DBEVT)
Specifies which database events are included in the definition. Additional information about the DBEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No database events are included in the definition.
- *ALL
- All database events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the database event identifier to be included.
Element 2: Event counter
- *NONE
- No database events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Resources affinity events (RSCAFNEVT)
Specifies which resources affinity events are included in the definition. Additional information about the RSCAFNEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No resources affinity events are included in the definition.
- *ALL
- All affinity events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the affinity event identifier to be included.
Element 2: Event counter
- *NONE
- No affinity events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Virtual I/O events (VRTIOEVT)
Specifies which virtual I/O events are included in the definition. Additional information about the VRTIOEVT options can be found in the Performance Management information at http://www.ibm.com/servers/eserver/iseries/perfmgmt/resource.html.
Note: This parameter is only valid if SLTEVT(*YES) is specified.
Single values
- *SAME
- The value does not change.
- *NONE
- No virtual I/O events are included in the definition.
- *ALL
- All virtual I/O events are included in the trace mode definition, or counted in the statistics mode definition.
Element 1: Event identifier
- event-identifier
- Specify the virtual I/O event identifier to be included.
Element 2: Event counter
- *NONE
- No virtual I/O events are counted.
- event-counter
- If TYPE(*STATS), events can be counted in one of eight counters (1,2,3,4,5,6,7 or 8). The event counter value is ignored if TYPE(*TRACE).
Element 3: Event format
The event format describes what data is collected for this event. *FORMAT1 provides the data used for most data analysis. The other formats allow for collection of other data related to these events.
- *FORMAT1
- The first event format is used to collect data for this event.
- *FORMAT2
- The second event format is used to collect data for this event.
- *FORMAT3
- The third event format is used to collect data for this event.
- *FORMAT4
- The fourth event format is used to collect data for this event.
Top |
Text 'description' (TEXT)
Specifies the text that briefly describes the Performance Explorer definition.
- *SAME
- The value does not change.
- *BLANK
- Text is not specified.
- 'description'
- Specify no more than 50 characters of text, enclosed in apostrophes.
Top |
Examples
Example 1: Using TYPE(*TRACE)
CHGPEXDFN DFN(TEST1) TYPE(*TRACE) JOB(*) MAXSTG(5000)
This command changes the performance definition named TEST1. When this definition is used to start a performance explorer session (STRPEX command), detailed trace information will be collected for the job that invoked the STRPEX command. A maximum of 5000 kilobytes of trace data will be collected. When the trace record storage area is full no more trace records will be collected.
Example 2: Using TYPE(*PROFILE)
CHGPEXDFN DFN(TEST2) TYPE(*PROFILE) PGM((MYLIB/MYSRVPGM1 *ALL *ALL *SRVPGM))
This command changes the performance explorer definition named TEST2. When this definition is used to start a performance explorer session (STRPEX command), performance profile information for service program MYSRVPGM1 in library MYLIB will be collected.
Top |
Top |