IBM Support

PI61683: WMQ JAVA THREADS FOR MQGET HANG AFTER WEBSPHERE MQ JMS MAINTENANCE IS APPLIED 16/05/04 PTF PECHANGE

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • Code introduced through PI46510/UI33362 can impact applications
    using an RRS enabled connection to WebSphere MQ ( including
    bindings mode connections from WAS ) When a connection to MQ is
    in an MQGET wait and either a message arrives or the wait
    interval expires, the code to post the waiting task can
    incorrectly determine that the task is no longer valid and skip
    the post call. This results in the connection remaining
    permanently suspended in the MQGET wait.
    .
    Additional keywords:
    WebSphere Application Server WAS WSAS
    IBM Integration Bus (formerly WebSphere Message Broker)
    

Local fix

  • n/a
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED: All users of WebSphere MQ for z/OS Version 7 *
    *                 Release 1 Modification 0                     *
    ****************************************************************
    * PROBLEM DESCRIPTION: The following two problems are fixed by *
    *                      this APAR.                              *
    *                      1:                                      *
    *                      After applying UI16412, RRS             *
    *                      applications issuing MQGET with         *
    *                      MQGMO_WAIT hang if RRS is restarted     *
    *                      while the application is waiting to be  *
    *                      posted.                                 *
    *                                                              *
    *                      2:                                      *
    *                      After applying UI33362, RRS             *
    *                      applications (including IIB and WAS)    *
    *                      issuing MQGET calls can hang            *
    *                      indefinitely if the MQGET is issued on  *
    *                      a different TCB to that originally used *
    *                      by a context, and the original TCB is   *
    *                      no longer active                        *
    ****************************************************************
    * RECOMMENDATION:                                              *
    ****************************************************************
    1:
    When RRS is restarted, any active RRS agents are driven through
    End of Task (EOT) processing. This should result in any such
    agents that are in an MQGET wait being posted when the handled
    is closed, however changes added to CSQM1P1W by PI08608 cause
    this post to be incorrectly skipped.
    
    2:
    When a getting application is being posted (for example, by the
    arrival of a suitable message, or the expiry of the get wait
    interval), code added to CSQM1P1W by PI46510 checks if the
    getter TCB is still active if there are any tasks in the getter
    address space in EOT processing.
    This check uses the tcbtoken stored in the getter's MTHR,
    however if the MQGET call was issued on a different TCB to the
    TCB the context originally used, the tcbtoken is incorrect and
    still relates to the original TCB. If that TCB has ended,
    CSQM1P1W incorrectly determines that the getter is going through
    EOT processing and skips posting the getter's ECB, leading to
    the reported problems.
    

Problem conclusion

  • CSQMGET is changed to ensure the MTHR associated with an RRS
    MQGET request is updated with the correct tcbtoken for the TCB
    currently being used for the MQGET request, allowing the check
    in CSQM1P1W to correctly determine if that TCB is still active
    or not.
    CSQM1P1W is changed to no longer skip the post when an RRS
    agent is driven through EOT.
    100Y
    CSQMALL
    CSQMGET
    CSQM1P1W
    

Temporary fix

  • *********
    * HIPER *
    *********
    

Comments

APAR Information

  • APAR number

    PI61683

  • Reported component name

    WMQ Z/OS V7

  • Reported component ID

    5655R3600

  • Reported release

    100

  • Status

    CLOSED PER

  • PE

    YesPE

  • HIPER

    YesHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2016-04-29

  • Closed date

    2016-05-12

  • Last modified date

    2016-08-17

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

    PI61141

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

    UI37799

Modules/Macros

  • CSQMALL  CSQMGET  CSQM1P1W
    

Fix information

  • Fixed component name

    WMQ Z/OS V7

  • Fixed component ID

    5655R3600

Applicable component levels

  • R100 PSY UI37799

       UP16/05/26 P F605 «

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.

[{"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Product":{"code":"SG19M","label":"APARs - z\/OS environment"},"Component":"","ARM Category":[],"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"7.1","Edition":"","Line of Business":{"code":"","label":""}}]

Document Information

Modified date:
17 August 2016