IBM Support

PH41372: STOR SHORTAGE WHEN LOADING MANY TABLES RAPIDLY INTO ACCELERATOR WHEN USING IBM INTEGRATED SYNCHRONIZATION FOR INCREMENTAL UPDATE

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • Various error conditions may occur when running many loads of
    replicated tables into an accelerator in rapid sequence when
    using IBM Integrated Synchronization.
    A rapid sequence of multiple table load operations into an
    accelerator is identified by multiple pairs of messages
    DSNI090I STARTING ASYNCHRONOUS LOG READER TASK and
    DSNI092I NORMAL TERMINATION OF ASYNCHRONOUS LOG READER TASK
    in a very short time period, typically less than 30 seconds.
    
    Storage related errors may then be seen, e.g. one of but not
    limited to:
    ABND=04E-00E2001F in DSNIDLGR.DSNSVBK +02DAE
    ABND=0C4-00000038 in DSNIDM.DSNIOBDJ +040B6
    MSGDSNI091I ERROR IN ASYNCHRONOUS LOG READER TASK RC8 RC00C9000A
    
    When loading a replicated table into an accelerator ABND04E
    RC00D34454 with LOC=DSNLXLLM.DSNLJMLR:0005 may occur and will
    cause replication to be paused.
    
    Additional keywords:
    IDAAV7R5/K DATAGATE/K
    MSGDSNI090I MSGDSNI092I
    

Local fix

  • n/a
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED:                                              *
    * All Db2 12 for z/OS users who either use                     *
    * IBM Db2 Analytics Accelerator with IBM                       *
    * Integrated Synchronization or IBM Db2 for                    *
    * z/OS Data Gate.                                              *
    ****************************************************************
    * PROBLEM DESCRIPTION:                                         *
    * Storage shortage when loading many                           *
    * tables in rapid sequence causing                             *
    * errors like but not limited to:                              *
    * ABEND0C4 RC00000038 DSNIDM.DSNIOBDJ                          *
    * OFFSET040B6, or ABEND04E RC00E2001F                          *
    * DSNIDLGR.DSNSVBK OFFSET02DAE. or                             *
    * MSGDSNI091I ERROR IN ASYNCHRONOUS LOG                        *
    * READER TASK RC8 RC00C9000A.                                  *
    * ABEND04E RC00D34454 with                                     *
    * LOC=DSNLXLLM.DSNLJMLR:0005 may occur                         *
    * when loading a replicated table and                          *
    * will cause replication to be paused.                         *
    ****************************************************************
    * RECOMMENDATION:                                              *
    * Apply corrective PTF when available                          *
    ****************************************************************
    When loading a large number of tables into an accelerator or
    data gate instance in a very rapid sequence and reading from
    the log is slow it is possible that storage allocated by
    log reading tasks accumulates until storage is exhausted
    causing various error situations. A rapid sequence of log
    reading tasks being started and stopped by the load processes
    is identified by many pairs of messages DSNI090I STARTING
    ASYNCHRONOUS LOG READER TASK and DSNI092I NORMAL TERMINATION OF
    ASYNCHRONOUS LOG READER TASK in a very short time period,
    typically more than 20 pairs in less than 30 seconds.
    When the termination of a log reading task during a table load
    lasts longer than 10s due to very low processing priority,
    ABEND04E RC00D34454 can be issued by DSNLJMLR:0005. This abend
    may cause replication to go into paused state.
    

Problem conclusion

  • The largest storage allocated by log reading tasks is
    deallocated while the termination request is processed to avoid
    accumulation of storage.
    When terminating a log reading task takes longer, any cleanup
    activity done by the watchdog process in parallel no longer
    causes a diagnostic abend in the termination routine.
    

Temporary fix

Comments

APAR Information

  • APAR number

    PH41372

  • Reported component name

    DB2 OS/390 & Z/

  • Reported component ID

    5740XYR00

  • Reported release

    C10

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2021-10-15

  • Closed date

    2021-12-28

  • Last modified date

    2022-02-01

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

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

    UI78759

Modules/Macros

  • DSNILGRT DSNLJMCF DSNILGWD DSNLJMLR DSNIOBDJ
    

Fix information

  • Fixed component name

    DB2 OS/390 & Z/

  • Fixed component ID

    5740XYR00

Applicable component levels

  • RC10 PSY UI78759

       UP22/01/05 P F201

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","label":"Data and AI"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Product":{"code":"SSEPEK","label":"DB2 for z\/OS"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"12.0"}]

Document Information

Modified date:
02 February 2022