PM85710: CHIN ABEND 5C6-0BAD0002, ,U=RP00087 ,C=R3600.701.DMC CSQIRREQ,M=CSQGFRCV,LOC=CSQILPLM.CSQIRREQ+000007AE

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Abend5C6-0BAD0002 occurred during a put to topic where we got a
  pageset full condition. The try to put the message to DLQ ends
  because the same page set is used and we recieved again page set
  full.
  
  The problem occurs when CSQIRREQ is invoked to backout as a
  result of the failure during MQPUT to a topic.
  CSQIRREQ is invoked to backout the nested unit of work due to
  the failure, and reads the last log records, 006CE151BF1D.
  This is a compensating log-record, due to a previous CSQIRREQ
  issued by
  CSQIMPU2 when it failed to allocate the necessary pages for the
  message. As a compensating log-record is being processed,
  CSQIRREQ sets
  LRCLRH->LRHLINK to 006CE151BEFB, which comes from LRHUNDO (set
  by CSQIUNDO when performing the backout requested previously by
  CSQIMPU2).
  CSQJR103 follows the log-records backwards from the current
  

Local fix

• n/A
  

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: The queue manager abends with           *
  *                      5C6-0BAD0002 issued by CSQIRREQ.        *
  *                      If the channel initiator is issuing the *
  *                      put, it will issue abend 5C6-00E7014F.  *
  *                      The abend occurs following an attempt   *
  *                      to put to a topic where the associated  *
  *                      subscriber queues are on a page set     *
  *                      that is almost full.                    *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  A message is being put to a topic. When the put to one of the
  subcriber queues fails with return code 2 reason 2192, the put
  is rolled back by invoking CSQIUNDO. The put failed after
  inserting the message body on its page, when requesting a
  spine-page for the queue, there is no more space left. This
  causes the MQPUT to be rolled back, and compensating log
  records are written to the log.
  
  When the put to the topic is being rolled back as a result of
  this (due to the persistent message delivery being set to all
  (durable) subscribers), CSQIRREQ is invoked, which will roll
  back the unit of work to a save point. The log is read
  backwards using LRHLINK, and when encountering a compensating
  log record from the MQPUT, the LRHUNDO value is copied into the
  LRHLINK value, and the next record is read.
  CSQIUNDO does not store an exact RBA in the LRHUNDO value (it
  stores the RBA of the compensated record minus 1), thus the
  next log read fails unexpectedly, causing the 0BAD0002 abend to
  be issued, which is propagated.
  

Problem conclusion

• The code was changed to restore a correct LRHLINK value,
  allowing the roll back of the put to the topic to succeed.
  010Y
  100Y
  CSQIRREQ
  

Temporary fix

Comments

APAR Information

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

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

  UK94710 UK94711

Modules/Macros

• CSQIRREQ
  

Fix information

• Fixed component name

  WMQ Z/OS V7

• Fixed component ID

  5655R3600

Applicable component levels

• R010 PSY UK94710

     UP13/07/24 P F307

• R100 PSY UK94711

     UP13/07/24 P F307

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.