IBM Support

PI72917: SQL TIMEOUT WHILE UPDATING CHECKPOINTREPSITORY TABLE

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • When a server and application have a mix of transaction
    policies (e.g. one local, one global), customers may see
    occasional timeouts on updates to the CHECKPOINTREPOSITORY
    table.
    

Local fix

  • Set the application and server transaction policies to the same
    value.
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED:  All users of WebSphere Compute Grid         *
    ****************************************************************
    * PROBLEM DESCRIPTION: Running a mix of jobs with the          *
    *                      property                                *
    *                      "com.ibm.websphere.batch.transaction.po *
    *                      licy" set to LOCAL and other            *
    *                      transaction policies in the Compute     *
    *                      Grid batch container results in         *
    *                      intermittent deadlocks.                 *
    ****************************************************************
    * RECOMMENDATION:                                              *
    ****************************************************************
    When running a mix of jobs in which the job property
    com.ibm.websphere.batch.transaction.policy is defined to LOCAL
    and other transaction policies in a batch container,
    intermittent sql deadlocks and rollbacks may be experienced.
    The other policies that the property could be set to are GLOBAL
    and the default COMPAT mode. These deadlocks and subsequent
    rollbacks can be seen in the batch container runtime and cause
    the executing jobs to fail.
    

Problem conclusion

  • Due to an inefficiency in the management of local storage
    associated with batch runtime threads, the transaction
    behaviours for LOCAL, GLOBAL and the default COMPAT mode can
    get used inappropriatly when those workloads are run
    interspersed with each other in a common batch container.
    Thread local storage management was fixed to ensure that LOCAL
    mode behaviour occured only when LOCAL mode is specified and
    GLOBAL behaviour occurs only when GLOBAL mode is specified.
    

Temporary fix

Comments

APAR Information

  • APAR number

    PI72917

  • Reported component name

    WXD Z COMP GRID

  • Reported component ID

    5655V6201

  • Reported release

    800

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2016-11-29

  • Closed date

    2016-12-13

  • Last modified date

    2016-12-13

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

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

Fix information

  • Fixed component name

    WXD Z COMP GRID

  • Fixed component ID

    5655V6201

Applicable component levels

  • R800 PSY

       UP

[{"Business Unit":{"code":"BU053","label":"Cloud \u0026 Data Platform"},"Product":{"code":"SS8NUZ","label":"WebSphere Extended Deployment Compute Grid for z\/OS"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"8.0","Line of Business":{"code":"LOB36","label":"IBM Automation"}}]

Document Information

Modified date:
18 October 2021