PI10235: WMQ ABENDS DURING RECOVERY FOR CF STRUCTURE WITH 00C5105B ABN=5C6-00C5105B,C=R3600.710,LOC=CSQELPLM.CSQEDSS2+xxxxxxxx

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• MQ abends is during recovery for CF structure with 00C5105B.
  The problem occurs when the queue-manager is in the middle
  of connecting to an application structure when an ENF 35
  event occurs (i.e. IXCYEnfFunctionStrAvail due to another
  queue-manager disconnecting from the structure).
  As the structure is not allocated, a
  IXCYEnfFunctionStrAvail is received by the queue manager.
  This causes qmgr to connect to the structure, allocating it,
  and invoke CSQERAD2 for recovery.
  .
  This happens during the very early stages of queue manager
  start up, meaning that the normal CF manager initialisation
  hasn't yet completed.
  The CF thread blocks haven't been built yet,
  nor the unit of work descriptors required for recovery.
  This causes CSQERAD2 to recover the admin structure
  entries for qmgr incorrectly.
  The queue manager abends with 00C5103A due to outstanding
  DB2 recovery, which is not related to this failure.
  .
  At the next attempt of the start up, the qmgr connects to
  the admin structure and starts rebuilding the data for
  other queue managers in the QSG (Queue Sharing Group),
  however it does not attempt to rebuild it's own data
  again, because it was marked as completed during the
  first start up.
  .
  The qmgr abends during recovery for CF structure with
  00C5105B. This abends occurs every time an attempt is
  made to recover the space map, causing other qmgr in
  the QSG.
  .
  ABN=5C6-00C5105B,C=R3600.710.CFM -CSQEDSS2,
  M=CSQGFRCV,LOC=CSQELPLM.CSQEDSS2+xxxxxxxx
  .
  Symptom             Description
  -------             -----------
  PIDS/5655R3600      Program id: 5655R3600
  RIDS/CSQELPLM#L     Load module name: CSQELPLM
  RIDS/CSQEDSS2       Csect name: CSQEDSS2
  AB/S05C6            System abend code: 05C6
  PRCS/00C5105B       Abend reason code: 00C5105B
  REGS/C0BCC          Register/PSW difference for R0C:-0BCC
  RIDS/CSQGFRCV#R     Recovery routine csect name: CSQGFRCV
  .
  Date Assembled:          20111013
  Module Level:            13.03GA
  Subfunction:             CFM  CSQEDSS2SMDS SPACE
  .
  
  .
  Additional Syptoms:
  ==================
  IEA794I SVC DUMP HAS CAPTURED:
  DUMPID=004 REQUESTED BY JOB (XXXMSTR)
  DUMP TITLE=XXX,ABN=5C6-00C5105B,U=SYSOPR  ,C=R3600.710.CFM
       -CSQEDSS2,M=CSQGFRCV,LOC=CSQELPLM.CSQEDSS2+xxxxxxxx
  CSQE131I !XXXX CSQERCF2 Recovery of structure <structure_name>
            completed
  
  The above abend does not occur if SMDS is not used. The
  following symptom can be observed without SMDS:
  Messages are unexpectedly recovered, or messages that are
  expected to be recovered are not restored as expected.
  .
  Additional keywords:
  MQSMDS/K
  

Local fix

• .
  none
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of WebSphere MQ for z/OS Version 7 *
  *                 Release 1 Modification 0.                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: During queue manager restart, the       *
  *                      CSQ_ADMIN structure is allocated,       *
  *                      indicated by message IXL015I in the     *
  *                      joblog, and message CSQE018I is issued  *
  *                      by module CSQERAD2.                     *
  *                      When recovering an application          *
  *                      structure, the queue manager abends     *
  *                      with 5C6-00C5105B.                      *
  *                      Any of the queue managers in the QSG    *
  *                      can subsequently abend with             *
  *                      5C6-00C5105B when recovering the same   *
  *                      application structure(s).               *
  *                      If SMDS is not used, the symptoms       *
  *                      include:                                *
  *                      - Consumed messages are being           *
  *                        recovered in error                    *
  *                      - Unconsumed messages are not being     *
  *                        recovered                             *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When CSQERAD2 is recovering the CSQ_ADMIN structure during
  startup in response to a ENF 35 event, the shared unit of work
  information is not correctly rebuilt for the queue manager
  that is recovering the admin structure.
  Thus when recovering application structures, some of the shared
  message operations are not replayed correctly, causing messages
  to be read back into the structure, unconsumed messages not to
  be correctly recovered, and in the case of SMDS, a space map
  issue in the SMDS data set for one or more queue managers.
  

Problem conclusion

• The startup recovery code was changed to not listen to ENF 35
  events, ensuring that recovery is processed by the correct
  code path.
  100Y
  CSQECONN
  CSQERAD2
  

Temporary fix

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

Comments

APAR Information

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

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

  UI15080

Modules/Macros

• CSQECONN CSQERAD2
  

Fix information

• Fixed component name

  WMQ Z/OS V7

• Fixed component ID

  5655R3600

Applicable component levels

• R100 PSY UI15080

     UP14/03/04 P F403 Ž

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.