IT19783: IBM MQ 8.0.0.6 queue manager terminates due to lack of stack space

APAR status

• Closed as program error.

Error description

• A queue manager terminated due to a lack of stack space. The
  queue manager should not have stopped automatically because of
  this lack of stack space - the intent is to catch that issue
  before it becomes a serious problem.
  
  In addition to the queue manager stopping, an FFST diagnostic
  file will be generated with the following characteristics:
  
  LVLS              :- 8.0.0.6
  Probe Id          :- KN378077
  Component         :- ksiEvalInternal
  Major Errorcode   :- krcE_RESOURCE_PROBLEM
  Minor Errorcode   :- OK
  Probe Description :- AMQ6125: An internal WebSphere MQ error has
  occurred.
  Comment1          :- Operating system stack limit reached,
  reduce selector complexity
  
  MQM Function Stack
  zlaMainThread
  zlaProcessMessage
  zlaProcessMQIRequest
  zlaMQGETM
  zsqMQGETM
  kpiMQGETM
  kqiWaitForMessage
  apiGetMessage
  aqmGetMessage
  aqhGetMessage
  aqhLocateMessage
  aqhCheckMsgChains
  ksiSqlTreeEvaluate
  ksiEval
  ksiEvalInternal
  ksiEvalInternal
  xcsFFST
  

Local fix

• Either increase the resources available to the queue manager, or
  reduce the workload.
  
  In the original scenario where this issue was observed, the
  cause of the high stack space usage was very complex message
  selector string. In this case, reduce selector complexity to
  avoid excessive amounts of stack space being used.
  

Problem summary

• ****************************************************************
  USERS AFFECTED:
  Systems running MQ version 7.1.0.8, 7.5.0.7 or 8.0.0.6 that are
  resource limited.
  
  
  Platforms affected:
  MultiPlatform
  
  ****************************************************************
  PROBLEM DESCRIPTION:
  When the low resource diagnostic FFST code is executed in
  combination with low resource availability, the queue manager
  stops due to lack of available system stack space (this FFST
  diagnostic should normally avoid this happening).
  
  A variation of the problem may be observed users running
  7.1.0.8 or 7.5.0.7. This manifested itself as the amqzmuc0
  process becoming unresponsive whilst it contested a lock within
  the diagnostic code, in this situation xecL_W_LONG_LOCK_WAIT
  Failure Data Capture (FDC) recordss, together with
  xecL_W_SEM_OWNER_DIED FDC records s may appear from the amqzlaa0
  process.
  

Problem conclusion

• MQ has been modified to minimise use of stack space whilst
  executing diagnostic FFST code at version 8, and prevent the
  diagnostic lock contention at v7.
  
  An improvement to the diagnostic routines made as part of this
  change has been delivered to MQ version 9 releases, hence the
  additional targets listed below.
  
  ---------------------------------------------------------------
  The fix is targeted for delivery in the following PTFs:
  
  Version    Maintenance Level
  v7.1       7.1.0.9
  v7.5       7.5.0.8
  v8.0       8.0.0.7
  v9.0 CD    9.0.4
  v9.0 LTS   9.0.0.2
  
  The latest available maintenance can be obtained from
  'WebSphere MQ Recommended Fixes'
  http://www-1.ibm.com/support/docview.wss?rs=171&uid=swg27006037
  
  If the maintenance level is not yet available information on
  its planned availability can be found in 'WebSphere MQ
  Planned Maintenance Release Dates'
  http://www-1.ibm.com/support/docview.wss?rs=171&uid=swg27006309
  ---------------------------------------------------------------
  

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:

Fix information

• Fixed component name

  WMQ BASE MULTIP

• Fixed component ID

  5724H7251

Applicable component levels

• R800 PSY

     UP