A fix is available
APAR status
Closed as new function.
Error description
This APAR enables IPR Unload utility to support a new report titled "Largest Database Records" and reduce the CPU time in producing the existing statistics reports printed in the HPSOUT data set.
Local fix
Problem summary
**************************************************************** * USERS AFFECTED: All users of IMS High Performance Unload * * for OS/390, Release 1 (FMID=H1IN110) * * who are using IPR Unload utility of * * IMS Parallel Reorganization for z/OS, * * Version 3 Release 1 (FMID=H25N310) * **************************************************************** * PROBLEM DESCRIPTION: This APAR provides code changes that * * are required to support new functions * * of IPR Unload utility. * **************************************************************** * RECOMMENDATION: Apply the maintenance for this APAR. * **************************************************************** This APAR enables IPR Unload utility to support a new report titled "Largest Database Records" and reduce the CPU time in producing the existing statistics reports printed in the HPSOUT data set. For the details of the new report, see the documentation change below.
Problem conclusion
Temporary fix
Comments
======================== = DOCUMENTATION CHANGE = ======================== IMS Parallel Reorganization for z/OS, V3R1, User's Guide (SC18922800) should be changed as follows: ================================================================ 3.0 The IPR Unload Utility 3.3 Restrictions : The IPR Unload utility supports PHDAM or PHIDAM partitions that are defined as capable of HALDB Online Reorganization(OLR) If one or more partitions are in the following HALDB OLR status, the IPR Unload utility cannot process the HALDB: - HALDB OLR is currently running. - HALDB OLR is stopped prior to the completion for the partition and one of the following options is specified: - DECOMPESS=NO - User exit - *CS format for PHDAM - *A1 format When none of the above options is specified, the IPR Unload utility can process the partition. The following options, however, are ignored: - BYINDEX=YES - COMPAREDLI=YES - DBSTATS=YES - KEYCHECK=ABEND/GG/GX - SKIPERROR=nnnnn and the following reports are suppressed: - Segment Statistics Report - Distribution of Database Record Sizes Report |- Largest Database Records Report - HDAM Root Addressable Area Statistics Report In this case, the performance decreases because the HSSR Engine passes the DL/I calls to the IMS's DL/I call handler from the partition. We recommend that you run the IPR Unload utility after the completion of the HALDB OLR. ================================================================ 3.8 Control Statements 3.8.1 Overview of Control Statements : The following control statements are provided to control the unload function: - DBRECDIST | - DBRECKEYS - DECODEDBD : ================================================================ 3.8.3 Control Statements for Unload Function (In Table 12, a new control statement DBRECKEYS is added.) Table 12. Control Statements for Unload Function. |------------------------------------------------------------- |Keyword: DBRECKEYS |Acceptable Values: 0-999 |Default: 10 |Description: Specifies the number of the largest | database records to be listed in | the Largest Database Records report. | See "DBRECKEYS Control Statement." |------------------------------------------------------------- ================================================================ (This is a new section, which is inserted next to "3.8.3.1 DBRECDIST Control Statement." ) |3.8.3.X DBRECKEYS Control Statement | |The DBRECKEYS control statement specifies the number of |the largest database records in the database to be listed |in the Largest Database Records report. | |This control statement is optional. If you do not explicitly |specify this control statement, DBRECKEYS=10 will be used. | | | +--10------+ | | | | >>___DBRECKEYS=--+--nnn-----+ | | nnn | Specifies a left-justified decimal number from 0 to 999. | The largest nnn database records in the database are listed | in the Largest Database Records report. The default is 10. | When DBRECKEYS=0 is specified, the report is suppressed. ================================================================ 3.8.3.11 STATREP Control Statement : (STATREP=)NO The following reports are not printed. CPU time is reduced by suppressing them. - Segment Statistics report - Distribution of Database Record Sizes report | - Largest Dartabase Record report - Data Set Statistics report - HDAM Root Addressable Area Statistics report - HISAM Statistics report Even if DBSTATS=YES is specified, the following reports are also suppressed: - HDAM RAP Chain I/O Summary report - DB Statistics report : ================================================================ 3.9 Reports and Statistics : The following reports are written in the HPSOUT data set: Reports that are generated for all database organizations: - Environment report - HPSIN Control Statements report - Decoded DBD report (if DECODEDBD=YES) - DBD Summary report - Segment Count Summary report - Segment Statistics report (3) | - Largest Database Recoeds report (3) - Data Set Statistics report (3) : (3) For HALDBs, the report is generated for each partition that is processed. When the high key selection is used, the report is generated in the order of the high key, and when the partition selection exit is used, it is generated in the order of the partition name. ================================================================ (This is a new section, which is inserted next to "3.9.10 Distribution of Database Record Sizes Report." ) |3.9.X Largest Database Records Report | |Figure XX is an example of a Largest Database Records report. |This report shows the largest database records in the database |or each HALDB partition. The number of the listed records can |be specified by the DBRECKEYS control statement. | |Figure XX. Largest Database Records Report | (Here, the report example is omitted.) | |The meaning of the fields in this report is as follows: | | PARTITION | The partition name. This field is printed only for a HALDB. | | RECORD SIZE | The database record size including the sizes of the segment | prefix portion and the segment data portion for each segment | | - For a variable-length segment (including the compressed | fixed-length segment), this includes the bytes padded by IM | - For a split segment, this includes a total of 6 bytes | for the address of the split destination segment, | the segment code, and delete byte. | This does not include the size of the virtual LPCK part. | | NBR OF SEGMENTS | The number of segment occurences in the database record. | | ROOT RBA | The RBA of the root segment. | If the database is HISAM or SHISAM, this field shows blank. | | ROOT KEY | The high 16 bytes of the root key value. The prefix C | indicates that it is a character value. Otherwise, | the values are shown in hexdecimal representation | with the prefix X. ================================================================ C.0 Appendix C. Database Tuning Scenarios C.1 Summary of Reports of the IPR Unload and the IPR Reload Utilities and Database Tuning : Reports used for database tuning : - Distribution of Database Record Sizes report This report can be used to check the average length of database records and the distribution of the length of database records to determine the block size or the CI size of database data sets. |- Largest Database Records report (IPR Unload utility only) | This report can be used to know the longest data base | records in the database. : ================================================================ C.2.4 Is the Problem with Long Database Records Solved Appropriately? : Whether the database record length is high can be determined by looking at the Distribution of Database Record Sizes report; this provides the average database record length and a table with the distribution of the database record length. |Also, you can see which the database records are long by lookin |at the Longest Database Records report. : ===================== End of Doc-Change ========================
APAR Information
APAR number
PK04792
Reported component name
IMS HIGHPERF UN
Reported component ID
5655E0600
Reported release
110
Status
CLOSED UR1
PE
NoPE
HIPER
NoHIPER
Special Attention
NoSpecatt / Xsystem
Submitted date
2005-04-25
Closed date
2005-05-30
Last modified date
2005-07-01
APAR is sysrouted FROM one or more of the following:
APAR is sysrouted TO one or more of the following:
UK03949
Modules/Macros
FABHB20 FABHB30 FABHC00 FABHDATE FABHFPH FABHFPH8 FABHFPH9 FABHFSUI FABHFSU7 FABHFSU8 FABHFSU9 FABHGDL7 FABHGDL8 FABHGDL9 FABHGND FABHGSA FABHLINK FABHPSL FABHPSL8 FABHPSL9 FABHSDB FABHSDB0 FABHTCB FABHTCBU FABHURG7 FABHURG8 FABHURG9 FABH001 FABH002 FABH0107 FABH0108 FABH0109 FABH0207 FABH0208 FABH0209 FABH0307 FABH0308 FABH0309 FABH040 FABH0427 FABH0428 FABH0429 FABH050 FABH060 FABH090 FABH093 FABH100 FABH105 FABH110 FABH1207 FABH1208 FABH1209 FABH2007 FABH2008 FABH2009 HPSSSCA7 HPSSSCA8 HPSSSCA9 HPSUHTBL HPSUPTRW HPSURCBM HPSURCTA HPSURDBD HPSURENV HPSURPAR HPSURPTH HPSURQAN HPSURR13 HPSUSTMT HPSUURG7 HPSUURG8 HPSUURG9 HPSUUXAI HPSU015 HPSU015A H1IN110J
Fix information
Fixed component name
IMS HIGHPERF UN
Fixed component ID
5655E0600
Applicable component levels
R110 PSY UK03949
UP05/06/02 P F506
Fix is available
Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.
[{"Line of Business":{"code":"LOB10"},"Business Unit":{"code":"BU029","label":"Software"},"Product":{"code":"SSCX89K","label":"IMS Tools"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"110"}]
Document Information
Modified date:
20 October 2020