PM92995: MESSAGE LOST AFTER ABEND0D6 DUE TO RRS FAILURE IF BROWSE WITH MQGMO_MSG_LOCK FOLLOWED BY MQGET

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• After RRS failure, ABEND0D6 occurs at CSQIRRSI which is trying
  to issue an Express_UR_Interest.
  The recovery processing does not undo the MQGET process, but it
  returns to the application with MQRC 2195.
  The application then issues commit which results in message
  lost. In this situation the message should have been backed out
  before returning the application with MQRC 2195.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of WebSphere MQ for z/OS Version 7 *
  *                 Release 0 Modification 1 and Release 1       *
  *                 Modification 0.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: An underlying abend occurs during the   *
  *                      get processing, such as, but not        *
  *                      limited to, 0D6-00000027 (RRS abend),   *
  *                      causing:                                *
  *                      a MQGET with MQGMO_LOCK for a shared    *
  *                      queue to complete with MQRC 2195        *
  *                      (MQRC_UNEXPECTED_ERROR) and the message *
  *                      to be removed from the queue.           *
  *                      a MQGET with MQGMO_SYNCPOINT for a      *
  *                      locked message on a shared queue to     *
  *                      complete with MQRC 2195                 *
  *                      (MQRC_UNEXPECTED_ERROR) and the message *
  *                      to be removed from the queue.           *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  An MQGET is being processed. As this is for a shared queue, the
  ETRQS and ETROP control blocks have been created.
  After the underlying abend occurred, the recovery processing
  that takes place in CSQIMGES does not undo the creation of the
  ETRQS and ETROP blocks, nor does it undo the destructive MQGET
  and put the message back into the browse locked state it was in
  before the destructive MQGET was issued.
  In the case of the browse locked message, the browse lock is not
  undone correctly, and the message remains unavailable to other
  getters.
  

Problem conclusion

• The code has been changed to correctly undo the browse lock if
  an error occurs during the browse + lock operation.
  The code has been changed to correctly undo the destructive get
  and restore the browse lock if an error occurs during the MQGET
  of a browse locked message.
  010Y
  100Y
  CSQESYNC
  CSQESYN1
  CSQIMGES
  

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:

  UK96902 UK96903

Modules/Macros

• CSQESYNC CSQESYN1 CSQIMGES
  

Fix information

• Fixed component name

  WMQ Z/OS V7

• Fixed component ID

  5655R3600

Applicable component levels

• R010 PSY UK96902

     UP13/09/27 P F309 Ž

• R100 PSY UK96903

     UP13/09/27 P F309 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.