IBM Support

PI27603: WMQ V8 ON Z/OS: ERROR CSQX526E (MESSAGE SEQUENCE ERROR) ON A SHARED RECEIVER CHANNEL DURING DISASTER RECOVERY TEST

A fix is available

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as program error.

Error description

  • In the test of disaster recovery cust performed the following
    :
          .
    1/ stop of all qmgrs of the MQ plex in all LPARs
    2/ start of one MSTR on the recovery site at
    3/ perform the recover CFSTRUCT of the structures which comprise
       the structure containing the SYSTEM.QSG.CHANNEL.SYNCQ
    4/ start of the CHIN
    5/ error CSQX526E on a shared receiver channel
    .
    The problem will occur if a queue manager is removed from
    the queue-sharing group and then added back prior to recovering
    the application structure.
    

Local fix

  • n/a
    

Problem summary

  • ****************************************************************
    * USERS AFFECTED: All users of WebSphere MQ Version 8          *
    *                 Release 0 Modification 0.                    *
    ****************************************************************
    * PROBLEM DESCRIPTION: When using CSQ5PQSG FORCE QMGR command  *
    *                      to remove a queue manager, it does not  *
    *                      issue any warnings when the queue       *
    *                      manager logs or SMDS are required to    *
    *                      recover CF structures in case of        *
    *                      failure.                                *
    *                                                              *
    *                      Also the RECOVER CFSTRUCT command or    *
    *                      auto recovery completes successfully    *
    *                      after it has only partially recovered   *
    *                      the data without issuing any error or   *
    *                      warning message when a queue manager    *
    *                      has been force removed as indicated     *
    *                      above.                                  *
    *                                                              *
    *                      The RECOVER CFSTRUCT command or auto    *
    *                      recovery can generate abend 0C4 in      *
    *                      CSQERRPB followed by abend 5C6-00E20014 *
    *                      in CSQSGMN that causes the queue        *
    *                      manager to terminate with reason        *
    *                      00C510AB.                               *
    ****************************************************************
    * RECOMMENDATION:                                              *
    ****************************************************************
    CSQ5PQSG REMOVE QMGR command checks if the logs or the SMDS
    datasets are required to recover structures and prevents the
    queue manager from being removed from the QSG after issuing
    error messages CSQE556E and/or CSQE558E. The CSQ5PQSG FORCE QMGR
    does not check and it does not issue any warning messages.
    
    The RECOVER CFSTRUCT command or the auto recover skip any logs
    required to recover a structure without issuing a warning when
    the logs belong to a queue manager that has either been force
    removed, or has been force removed and added back into the QSG
    and hasn't been restarted. The recover command seems to complete
    normally but the data has been partially recovered.
    
    If during RECOVER CFSTRUCT command or auto recovery, DB2 becomes
    unavailable or there is loss of connectivity to the CF structure
    abend 0C4 can be generated in CSQERRPB followed by 5C6-00E20014
    in CSQSGMN resulting in queue manager termination with reason
    00C510AB.
    

Problem conclusion

  • CSQ5PQSG FORCE QMGR command will issue new warning messages
    CSQU586I and/or CSQU587I when the queue manager logs or SMDS
    are required to recover a CF structure.
    
    RECOVER CFSTRUCT or auto recovery will fail if there are logs
    required to recover the structure but they cannot be accessed
    due to the queue manager having been FORCE removed from the QSG.
    New error messages CSQE161E and CSQE112E are issued to report
    this condition.
    
    RECOVER CFSTRUCT or auto recovery has checks added to avoid
    generating abend 0C4 trying to access unallocated storage.
    
    ========= DOC Changes for V800 Knowledge Center ===============
    Add the following page to the Knowledge Center for MQ V800:
    
    WebSphere MQ
    > Reference
        > Diagnostic messages
            >  Messages for WebSphere MQ for z/OS
                > Messages
                   > Coupling Facility manager messages (CSQE...)
    
    CSQE161E: csect-name QSG state is inconsistent,
    no XCF data for queue manager qmgr-number
    
    Explanation
    
    A RECOVER CFSTRUCT command or automatic structure recovery was
    unable to read all the log data required for recovery as there
    was no XCF data for one of the queue managers in the QSG.
    qmgr_number is the number of the affected queue manager in the
    MQ DB2 tables.
    
    System action
    
    Processing of the command is terminated.
    Automatic recovery of the structure will not be attempted.
    
    System programmer response
    
    If the queue manager with number qmgr-number in the MQ DB2
    tables has been force removed from the QSG and added back into
    the QSG, start the queue manager and issue the RECOVER CFSTRUCT
    command again.
    
    Otherwise, reset the structure to an empty state by issuing the
    RECOVER CFSTRUCT TYPE(PURGE) command.
    
    ==============================================================
    
    Add the following page to the Knowledge Center for MQ V800:
    
    WebSphere MQ
    > Reference
        > Diagnostic messages
            >  Messages for WebSphere MQ for z/OS
                > Messages
                  > Utilities messages (CSQU...)
    
    CSQU586I: QMGR qmgr-name entry being removed from QSG qsg-name,
    needed for structure struc-name backup
    
    Explanation
    
    The queue manager named by qmgr-name is being force removed from
    the queue-sharing group named by qsg-name and it has information
    about backups for structure struc-name. (The value shown for
    struc-name is the 12-character name as used by WebSphere MQ, not
    the external name used by z/OS which includes the queue-sharing
    group name.)
    
    If the queue manager qmgr-name is added back to the
    queue-sharing group it will cause an inconsistent state that
    could prevent structure struct-name from being recovered should
    it fail before a structure backup is taken.
    
    If the queue manager is needed for more than one structure,
    this message will be issued for each one.
    
    System action
    
    Queue manager is removed from queue-sharing group and the
    utility program ends with return code 4.
    
    System programmer response
    
    If CF structure struc-name is usable, take a backup of CF
    structure struc-name as soon as possible using another queue
    manager in the queue-sharing group.
    
    Otherwise, if the queue manager qmgr-name is added back to the
    queue-sharing group it should be restarted before recovering
    structure struc-name.
    
    ==============================================================
    
    Add the following page to the Knowledge Center for MQ V800:
    
    WebSphere MQ
    > Reference
        > Diagnostic messages
            >  Messages for WebSphere MQ for z/OS
                > Messages
                   > Utilities messages (CSQU...)
    
    CSQU587I: QMGR qmgr-name entry being removed from QSG qsg-name,
    SMDS for structure struc-name is not empty
    
    Explanation
    
    The queue manager named by qmgr-name is being removed from the
    queue-sharing group named by qsg-name while it owns a shared
    message data set for structure struc-name which is not marked as
    empty, so it may still contain current message data. (The value
    shown for struc-name is the 12-character name as used by
    WebSphere MQ, not the external name used by z/OS which includes
    the queue-sharing group name.)
    
    System action
    
    Queue manager is removed from queue-sharing group and the
    utility program ends with return code 4.
    
    Messages on the SMDS for queue manager qmgr-name will remain
    accessible as long as the SMDS is retained.
    

Temporary fix

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

Comments

APAR Information

  • APAR number

    PI27603

  • Reported component name

    WMQ Z/OS 8

  • Reported component ID

    5655W9700

  • Reported release

    000

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    YesHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2014-10-14

  • Closed date

    2014-12-17

  • Last modified date

    2015-02-03

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

    PI21137

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

    UI23965 UI23966 UI23967 UI23968 UI23969 UI23970

Modules/Macros

  • CSQERCFT CSQERCF1 CSQERRPB CSQFEDIC CSQFEDIE CSQFEDIF CSQFEDIK
    CSQFEDIU CSQFMTXC CSQFMTXE CSQFMTXF CSQFMTXK CSQFMTXU CSQFSTXC
    CSQFSTXE CSQFSTXF CSQFSTXK CSQFSTXU CSQ5PQSG
    

Fix information

  • Fixed component name

    WMQ Z/OS 8

  • Fixed component ID

    5655W9700

Applicable component levels

  • R000 PSY UI23965

       UP15/01/29 P F501 ¢

  • R001 PSY UI23966

       UP15/01/29 P F501 ¢

  • R002 PSY UI23967

       UP15/01/29 P F501 ¢

  • R003 PSY UI23968

       UP15/01/29 P F501 ¢

  • R004 PSY UI23969

       UP15/01/29 P F501 ¢

  • R005 PSY UI23970

       UP15/01/29 P F501 ¢

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":"BU053","label":"Cloud & Data Platform"},"Product":{"code":"SSYHRD","label":"IBM MQ"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Line of Business":{"code":"LOB45","label":"Automation"}}]

Document Information

Modified date:
01 May 2020