PQ83386: IMSV9 HALDB ONLINE REORGANIZATION SUPPORT FOR IPR V2

Obtain the fix for this APAR.

APAR status

• Closed as new function.

Error description

• IMSV9 HALDB Online Reorganization Support for IPR V2
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of IPR Driver of IMS Parallel      *
  *                 Reorganization for z/OS Version 2 Release 1  *
  *                 (FMID=H25N210) who are to run IPR Driver or  *
  *                 IPR Unload utility on IMS Version 9 (IMS V9) *
  ****************************************************************
  * PROBLEM DESCRIPTION: By this maintenance, IMS Parallel       *
  *                      Reorganization for z/OS Version 2       *
  *                      Release 1 is enabled to run on IMS      *
  *                      V9 for online reorganization capable    *
  *                      (OLR-capable) HALDB partitions in which *
  *                      the M-through-V and Y data sets are     *
  *                      active. This maintenance also provides  *
  *                      the support for the new optional user   *
  *                      module DFSIDEF0 for IMS V9, which       *
  *                      contains the batch and utilities        *
  *                      installation default for DBRC.          *
  ****************************************************************
  * RECOMMENDATION: Apply the maintenance for this APAR.         *
  ****************************************************************
  By the maintenance provided by this APAR, IPR V2 is enabled to
  run on IMS V9 for OLR-capable HALDB partitions in which
  M-through-V and Y data sets are active. IPR Driver now supports
  the reorganization of OLR-capable HALDB partitions whether the
  A-through-J and X or the M-through-V and Y data sets are active.
  IPR Unload utility now supports processing of OLR-capable HALDB
  partitions whether the A-through-J and X or the M-through-V and
  Y data sets are active. IPR Unload utility also supports
  OLR-capable HALDB partitions in which online reorganization is
  suspended. New optional user module DFSIDEF0 for IMS V9, which
  contains the batch and utilities installation default for DBRC,
  is supported for IPR Driver, IPR Unload, IPR Reload, and and IPR
  DB Scan utilities.
  

Problem conclusion

Temporary fix

Comments

• The reloading of a HALDB partition by IPR Driver is always into
  the A-through-J and X data sets as IMS HD Reorganization Reload
  does. For details, see the following documentation change.
  <<< DOCUMENTATION CHANGE  >>>
     IMS Parallel Reorganization for z/OS Version 2 Release 1
  User's Guide (SC27-1180-02) should be modified as follows.
    * * *    * * *    * * *    * * *    * * *    * * *    * * *
    Chapter 1. Introduction
    1.2.1 Software Requirements
     IPR V2 requires the following products and maintenance:
  |  10. If you are to reorganize HALDB partitions in which the
  |   M-through-V and Y data sets are active , the following
  |   maintenance must be applied:
  |    - The maintenance provided by APAR PQ83387 for IMS
  |      High Performance Unload for OS/390 Release 1
  |    - The maintenance provided by APAR PQ85616 for IMS
  |      High Performance Load for OS/390 Release 1
    Chapter 2. IMS Parallel Reorganization Driver
    2.1.2.5 Automated Data Set Name Swapping
     Notes:
  |  3. If you specify NAMESWAP=YES for HALDB and the unloading
  |   and reloading process completed normally, the automated
  |   name swapping is always performed even if the database has
  |   logical relationship.
  |  4. If you reorganize an OLR-capable HALDB partition, the
  |   shadow data set names are always swapped with the
  |   A-through-J and Y data set names.
    2.2 System Flow of IPR Driver
  |  Note: The shadow data set names are always swapped with
  |   the A-through-J and Y data set names if an OLR-capable
  |   HALDB partition is reorganized.
    2.6.3.1 Database and Primary Index
  #     In this section, the following phrases should be
  #  changed "shadow data sets":
  #  - shadow output data set(s)   -> shadow data set(s)
  #  - output shadow data set(s)   -> shadow data set(s)
  #  - output database(s)          -> shadow database(s)
  #  - output database data set(s) -> shadow database data set(s)
  #  - output partition(s)         -> shadow partition(s)
  #  - output ILDS                 -> shadow ILDS
  #     The term "output data sets" should be understood as the
  #  data sets that you get as the reorganized database or index
  #  data sets. The term "shadow data sets" should be understood
  #  as the data sets in which the reload task loads segments or
  #  the index builder loads index records.
  #     The last paragraph of this section should be changed as
  #  as follows:
  |     On how the DD names for shadow data sets for non-HALDB are
  |  determined, see "DDNMAP Control Statement" in IPR V2 User's
  |  Guide. The DD names for shadow data sets for a HALDB
  |  partition are determined as shown in the following table.
  |  Note that, for the OLR-capable partition, the shadow DD names
  |  are determined depending on which side of partition data sets
  |  is active.
  |   .---------------------------------------------------------.
  |   |         |                DD name suffix                 |
  |   | Data    |-----------------------------------------------|
  |   | set     | When A-J and X active | When M-V and Y active |
  |   | type    |-----------------------------------------------|
  |   |         | Original  | Shadow    | Original  | Shadow    |
  |   |---------------------------------------------------------|
  |   | DSG A-J |    A-J    |    M-V    |    M-V    |    A-J    |
  |   | ILDS    |     L     |     Z     |     L     |     Z     |
  |   | PINDEX  |     X     |     Y     |     Y     |     X     |
  |   '---------------------------------------------------------'
  |    Note: DSG denotes the data set group, ILDS is the indirect
  |          list data set, and PINDEX is the primary index.
    2.6.3.3 Dynamic Allocation of Original and Shadow Data Sets
  |     The way in which original and shadow data sets for
     database and indexes are allocated is determined by the
     DYNALLOC control statement.
        By default, the data sets are allocated dynamically. For
  |  a non-partitioned database, the names for the shadow data
     sets are determined automatically, on the basis of data set
  |  names defined in the DFSMDA members for the original database
  |  and indexes. For PHDAM and PHIDAM, the names for the shadow
  |  data sets are determined, on the basis of the A-through-J and
  |  X data set names registered to RECON.
  |     For a non-partitioned database, the name of a shadow
     database data set consists of the name of the corresponding
     input data set followed by a suffix (the default is .Z). For
  |  example, suppose your original data set is named
  |  IMS.TESTDB01. By default, the shadow data set is named
  |  IMS.TESTDB01.Z. If you specify 'DYNALLOC=,A', the shadow data
  |  set is named IMS.TESTDB01.A. You can specify any alphabetic
     character except T as the suffix. The same rule also applies
  |  to the shadow primary index for a HIDAM database, a shadow
  |  secondary index, and a shadow indirect list data set.
  |     If the A-through-J and X data sets of a PHDAM or PHIDAM
  |  partition is active, the same naming rule as is applied to
  |  non-partitioned database applies. If the M-through-V and Y
  |  data sets of a PHDAM or PHIDAM partition is active, the
  |  shadow data set name for the partition is determined from the
  |  names of A-through-J and X data sets. For example, suppose
  |  your original data set name of DSG A is IMS.TESTDB01.M0001,
  |  the default shadow data set name is the corresponding A-side
  |  data set name IMS.TESTDB01.A0001 suffixed with .Z, that is,
  |  it is IMS.TESTDB01.A0001.Z. And, by the name swapping
  |  processing, the shadow data set is renamed to the A-side data
  |  set name IMS.TESTDB01.A0001. The original data set
  |  IMS.TESTDB01.M0001 is not renamed and remains as it is.
        For details, see "DYNALLOC Control Statement" and
  |  "NAMESWAP Control Statement".
    2.6.3.5 Data Set Names after Reorganization
        If you specify NAMESWAP=YES, ...
  |     For a non-partitioned database and a PHDAM and PHIDAM
  |  partition in which A-through-J and X data sets are active,
  |  each original data set name, including the VSAM component
  |  name, is renamed to the original name plus the suffix .T.
     The renamed original data sets are not deleted ...
  |     For a PHDAM and PHIDAM partition in which M-through-V and
  |  Y data sets are active, the original data sets are not
  |  renamed and remain as they are. In this case, the shadow data
  |  sets are renamed to the A-through-J and X data set names.
        You can see how the data sets were renamed by IPR
     Driver, by checking the Result of Name Swapping report.
    2.7.6.7 DYNALLOC Control Statement
        ...
  |     If DYNALLOC=(xxx,YES,xxx) is specified for non-HALDB, the
     data set names for ...
  |     If DYNALLOC=(xxx,YES,xxx) is specified for a HALDB
  |  partition the active DBDS of which is in A-through-J and X
  |  side, the shadow database data set names are created in the
  |  same way as that for non-HALDB. If DYNALLOC=(xxx,YES,xxx) is
  |  specified for a HALDB partition the active DBDS of which is
  |  in M-through-V and Y side, the data set names for the shadow
  |  database are created by adding a suffix character c to the
  |  name of the corresponding A-through-J and X data sets.
    2.7.6.12 NAMESWAP Control Statement
        The NAMESWAP control statement specifies...
        If you specify NAMESWAP=NO, the name swapping is not done.
     If you specify NAMESWAP=YES, the shadow data sets are
     renamed to the original data set names, that is, the sets of
  |  names of original and shadow data sets are swapped. For a
  |  PHDAM and PHIDAM partition the active DBDS of which is in
  |  M-through-V side, the shadow data sets are renamed to the
  |  data set names of A-through-J side, and the original data
  |  sets are not renamed. NAMESWAP=NO is the default.
        If you specify NAMESWAP=YES, the following conditions
     must be met: DBRC is active; all databases being reorganized,
  |  including the secondary indexes, are registered to DBRC; and,
  |  names that are used for the shadow data sets for the primary
     database and primary index are not registered to DBRC.
  |     If you specify NAMESWAP=YES, the names of the primary
  |  database and primary index data sets will be swapped. The
  |  names of the secondary index databases will be swapped
     only if IMS Index Builder task is activated and ...
  |     If you specify NAMESWAP=YES for a non-partitioned
  |  database and a PHDAM and PHIDAM partition the active DBDS of
  |  which is in A-through-J side, the original data set names are
     temporarily suffixed with .T. The original data set name,
     including the VSAM component names, must contain ...
        If you specify NAMESWAP=YES and the data set name
     swapping succeeds, Reorganization Manager issues DBRC
     NOTIFY.REORG commands for the reorganized database and the
  |  primary index. For each PHDAM and PHIDAM partition the active
  |  DBDS of which is in M-through-V and Y side, DBRC CHANGE.DB
  |  OLRDBDS(A) is issued. If you specify INDEXBLD=YES, the
     NOTIFY.REORG command is also issued for each secondary index.
    2.8.2.4 Data Set Information Report
  #     The following two lines are printed for each OLR-capable
  #  HALDB partition:
  #     ACTIVE DBDS BEFORE REORGANIZATION.... active_side
  #     ACTIVE DBDS AFTER  REORGANIZATION.... A-J note_text
  #  Here, active_side is either A-J or M-V. The note_text is
  #  "(NOTE: POST PROCESSING IS NECESSARY)" and it is printed when
  #  the automated post reorganization processing was not done.
  #     New report lines that shows the output data set names are
  #  printed if an OLR-capable HALDB partition is processed.
    2.9.3 Manual DBRC Notification Processing
         ...
        The way the DBRC notification processing is done varies at
     installations, but a typical sequence of steps is as follows:
     1. Notify DBRC that ...
  |        NOTIFY.REORG DBD(...) DDN(...) USID(usid) CURRENT
        Note: For a non-HALDB, it is assumed ...
  |     For a HALDB partition, if the active DBDS before the
  |     parallel reorganization was in the M-through-V and Y side,
  |     you need to issue the following command for the partition:
  |        CHANGE.DB DBD(part_name) OLRDBDS(A)
  |  2. Take an image copy...
  |        NOTIFY.UIC DBD(...) DDN(...) USID(usid) CURRENT
           ...
    Chapter 3 IPR Unload Utility
    3.3   Restrictions
             ...
  |     IPR Unload utility supports OLR-capable PHDAM or PHIDAM
  |  partitions. If OLR is in progress in a HALDB partition, the
  |  utility cannot process the HALDB. If OLR is suspended for a
  |  HALDB partition, the utility aborts if one of the following
  |  options is specified : DECOMPESS=NO, an user exit, the unload
  |  format *CS for PHDAM, the unload format *A1 for PHDAM and
  |  PHIDAM. If none of these options is specified, the utility
  |  can process the HALDB. But the following options are
  |  ignored: BYINDEX=YES, COMPAREDLI=YES, DBSTATS=YES,
  |  KEYCHECK=ABEND/GG/GX, and SKIPERROR=nnnnn, and the following
  |  reports are not printed: Segment Statistics Report,
  |  Distribution of Database Record Sizes Report, and HDAM Root
  |  Addressable Area Statistics Report. In this case, the
  |  performance can be degraded, IBM recommends that you run the
  |  IPR Unload utility after the OLR has been completed.
    3.8.3.3 DECOMPRESS Control Statement
     Notes:
  |  3. If DECOMPRESS=NO option is specified and one or more
  |  partitions of PHDAM or PHIDAM are in OLR cursor-active
  |  status, IPR Unload utility ends abnormally.
    3.8.3.5 OUTPUT Control Statement
     format
      Specifies one of the following unloaded data formats: ...
      Notes:
  |   3.  If *CS or *A0 format is specified and one or more
  |   partitions of PHDAM or PHIDAM are in OLR cursor-active
  |   status, IPR Unload utility ends abnormally.
        ...
     Notes: (In the last notes section)
  |  3. If an user exit routine is specified and one or more
  |  partitions of PHDAM or PHIDAM are in OLR cursor-active
  |  status, IPR Unload utility ends abnormally.
    3.8.3.6 Unloaded Data Formats
     Notes:
     2. RECA1RBA is the RBA of the segment prefix as returned
  |  by IMS. The RBA is the one that IMS uses internally. Thus,
  |  for an OSAM database the size of which exceeds 4 giga bytes
  |  or an OLR-capable HALDB partition, the RBA is not necessarily
  |  an actual relative byte address.
    3.8.4.4 BYINDEX Control Statement
     Notes:
  |  2. BYINDEX=YES is ignored if one or more partitions of PHDAM
  |  or PHIDAM are HALDB OLR cursor-active status.
    3.8.4.5 COMPAREDLI Control Statement
     Notes:
  |  3. COMPAREDLI=YES is ignored if one or more partitions of
  |  PHDAM or PHIDAM are in OLR cursor-active status.
    3.8.4.7 DBSTATS Control Statement
     Notes:
  |  2. DBSTATS=YES is ignored if one or more partitions of
  |  PHDAM or PHIDAM are HALDB OLR cursor-active status.
    3.8.4.9 KEYCHECK Control Statement
     Notes:
  |  2. KEYCHECK=ABEND/GG/GX is ignored if one or more partitions
  |  of PHDAM or PHIDAM are in OLR cursor-active status.
    3.8.4.12 SKIPERROR Control Statement
     Notes:
  |  2. SKIPERROR=nnnnnnnnn is ignored if one or more partitions
  |  of PHDAM or PHIDAM are in OLR cursor-active status.
    Appendix F. Messages and Codes
    F.3.4 HPSG Messages
  #  The following message added by APAR PQ80190 is modified:
  |  HPSG0401E ONLINE REORG ACTIVE FOR PARTITION=pppppppp
  |             (DBD=dddddddd)
  |  Explanation: The online reorganization (OLR) cursor of the
  |  partition ppppppp of the HALDB dddddddd is active. The
  |  Parallel Reorganization of the partition the OLR cursor of
  |  which is active is not supported.
  |  System Action: IPR Driver ends with a return code of 8.
  |  Programmer Response: Complete the online reorganization for
  |  the partition indicated in the message before you run the
  |  IPR Driver job. If you are to run the IPR Driver job for
  |  the multiple partitions, you need to check if the OLR
  |  cursors of all partitions are not active.
  #  The following messages are modified:
     HPSG0316I CHANGE.DB DBD(dbdname) parameter
     Explanation: The CHANGE.DB command has been successfully
     completed for the database or HALDB partition dbdname.
     parameter is one of the following:
  |   - OLRDBDS(A)
      - READON
      - NOAUTH
      - AUTH READOFF
     System Action: IPR Driver continues processing.
     Programmer Response: None.
     HPSG0316E CHANGE.DB DBD(dbdname) parameter --- ERROR (CC=xx)
     Explanation: An error occurred when IPR Driver issued the
     CHANGE.DB command for the database or a HALDB partition
     dbdname. The condition code returned from DSPURX00 is xx.
      parameter   Report of Reference
     -----------  ------------------------------------------------
  |  OLRDBDS(A)   DBRC AUTHORIZATION PROCESSING FOR NAME SWAPPING
     READON       DBRC AUTHORIZATION PROCESSING FOR REORGANIZATION
     NOAUTH       DBRC AUTHORIZATION PROCESSING FOR REORGANIZATION
                    or
                  DBRC AUTHORIZATION PROCESSING FOR NAME SWAPPING
     AUTH READOFF DBRC NOTIFICATION PROCESSING
     System Action: IPR Driver ends with a return code of 8.
     Programmer Response: Refer to the messages issued from the
     DBRC utility, and identify the cause of the error. The
     messages are printed in the HPSOUT2 data set.
    F.3.7 HPSU Messages
  |  HPSU3159E UNSUPPORTED OPTION FOR OLR ACTIVE PARTITIONS: text
  |  Explanation: The string "text" is one of the following:
  |  "DECOMPRESS=NO," "USEREXIT," "*CS FORMAT" (for PHDAM), or
  |  "*A1 FORMAT." These options cannot be used for a partition
  |  in which OLR-cursor is active.
  |  System Action: IPR Unload ends abnormally.
  |  Programmer Response: Remove this option or rerun the IPR
  |  Unload job after OLR for all the partitions are completed.
  |  HPSU3160W ONLINE REORG ACTIVE FOR PARTITION=pppppppp
  |            (DBD=dddddddd)
  |  Explanation: The HALDB Online Reorganization (OLR) processing
  |  for the partition pppppppp was stopped prior to completion
  |  and the OLR cursor is still active. HSSR Engine will process
  |  the partition which is comprised of both the A-through-J and
  |  X data sets as well as the M-through-V and Y data sets.
  |  System Action: IPR Unload continues processing. The
  |  following options are ignored: BYINDEX=YES, COMPAREDLI=YES,
  |  DBSTATS=YES, KEYCHECK=ABEND/GG/GX, and SKIPERROR=nnnnn.
  |  Programmer Response: Refer to additional message.
  

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

  UQ86560

Modules/Macros

• HPSGBBL6 HPSGBBL7 HPSGBBL8 HPSGBBL9 HPSGGDIB
  HPSGGDLB HPSGGIOB HPSGGIO0 HPSGGRMB HPSGGSTB HPSGMPA0 HPSGMSGT
  HPSGRA00 HPSGRA10 HPSGRA20 HPSGRA30 HPSGRDD0 HPSGRD00 HPSGRP00
  HPSGRP20 HPSGRS00 HPSGRS10 HPSGRS30 HPSGTM00 HPSGUBBL
  

Fix information

• Fixed component name

  IMS PARALLEL RE

• Fixed component ID

  5655F7400

Applicable component levels

• R210 PSY UQ86560

     UP04/03/26 P F403

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.