PH62197: IBM MQ Z/OS: A TIMING ISSUE WITH SHARED CHANNEL ADOPTION IN THE IGQ ENVIRONMENT

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Undelivered messages were put to dead-letter queue with
  CSQM064I after one of QSG members (CSQ1) restarted:
  15.18.49 S0683344  CSQM064I !CSQ3 CSQMIGQA Intra-group queuing
  agent put messages to dead-letter queue
  15.18.49 S0683342  CSQM064I !CSQ2 CSQMIGQA Intra-group queuing
  agent put messages to dead-letter queue
  
  In the scenario the following sequence of events leads to the
  reported problem:
  
  Channel INBOUND.SHARED.CHL1 is active on CSQ3 (started several
  days earlier)
  15:17:42 STOP QMGR MODE(FORCE) is issued on CSQ3
  15:17:43 The channel initiator begins termination processing
  The group listener is stopped
  Channel INBOUND.SHARED.CHL1 is stopped by chinit termination
  processing
  15.17.48 Channel INBOUND.SHARED.CHL1 starts on CSQ2 and begins
  adoption processing.
  A Temporary Dynamic Queue is created, and as CSQ3 is still
  running, a 'kill channel' message is put to CSQ3 via IGQ
  (however it won't be processed due to the in-progress STOP QMGR
  command)
  15.17.48 CSQ3CHIN stops
  15.18.58 Channel adoption fails on CSQ2, and the queue is
  deleted
  15.18.03 CSQ3MSTR stops
  
  Partly this symptom is exposed because the dynamic queue type
  is set to TEMPDYN, allowing the queue to be deleted before the
  restarting queue manager puts the reply (by default the model
  queue SYSTEM.DEFAULT.MODEL.QUEUE is defined as
  DEFTYPE(PERMDYN), in which case the queue would not have been
  deleted and the reply message would have simply sat on the
  reply queue instead of being put to the dead letter queue).
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of IBM MQ for z/OS Version 9       *
  *                 Release 2 Modification 0,                    *
  *                 Release 3 Modification 0 and                 *
  *                 Release 4 Modification 0.                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: When restarting a queue manager that is *
  *                      a member of a queue sharing group and   *
  *                      has a shared receiver channel running,  *
  *                      it is possible that a message is put to *
  *                      the dead letter queue of the queue      *
  *                      manager in the queue sharing group that *
  *                      tries to adopt the receiver channel.    *
  *                      This is displayed with log message      *
  *                      CSQM064I.                               *
  ****************************************************************
  The code which handles the adoption process for a shared channel
  sends a "kill channel" message from a queue sharing group member
  to the queue manager where the channel was running, if the
  channel is orphaned. The queue manager that is trying to adopt
  the channel only waits briefly for a reply from the
  other queue manager. If this queue manager is stopping or
  restarting, the message is delivered when it starts and it can
  then send a reply.
  
  If the reply to the 'Kill channel' message was sent to a
  temporary dynamic queue, this queue will be deleted past the
  waiting time from the adopting queue manager, leaving no
  destination queue for the reply. As a result of this, the reply
  is sent to the dead letter queue.
  

Problem conclusion

• The code has been changed to set a 10 seconds expire time to the
  'KillChl' reply message when performing adoption of a shared
  channel. As a result, the message will not be orphaned or sent
  to the dead letter queue.
  

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:

  UI99425 UI99426 UI99427

Modules/Macros

• CSQXRCSI
  

Fix information

• Fixed component name

  IBM MQ Z/OS V9

• Fixed component ID

  5655MQ900

Applicable component levels

• R200 PSY UI99427

     UP25/01/07 P F501

• R300 PSY UI99426

     UP25/01/07 P F501

• R400 PSY UI99425

     UP25/01/07 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.