IC82919: WEBSPHERE MQv5.3.1 FOR HP NONSTOP USING API EXITS REPORT PROBE ZF137003 AND AMQ9518: "AMQRFCDA.DAT not found"

WebSphere MQ for HP NonStop Server V5.3.1 fix pack 5.3.1.9

APAR status

• Closed as program error.

Error description

• Queue manager (receiver channel side) with API exits generate
  FDCs containing probe ZF137003 and error log entries for AMQ9518
  in the form of -  File '/mqs/WMQtest/var/mqm/@ipcc/AMQRFCDA.DAT'
  not found.
  
  Channels go into 'retry' state and FDC's are generated reporting
  symptoms similar to the ones below:
  
  Sequence of error:
   AMQ6118 Internal WMQ error has occured (program:MQECSVR)
   AMQ6184 Internal WMQ error has occured on qmgr QM001
                                             (program:MQECSVR)
   AMQ9518 File '/mqs/WMQtest/var/mqm/@ipcc/AMQRFCDA.DAT' not
  found
   AMQ9999 Channel program ended abnormally
  
  FDC generated by amqrmppa_r process:
  LVLS              :- 531.8  FP5318
  Probe Id          :- ZF137003
  Component         :- zfcApiExitReadCB
  Program Name      :- /mqs/WMQtest/opt/mqm/bin/amqrmppa_r
  Process           :- 2,1129 $X0B5X OSS(266993757)
  Thread            :- 00000083
  Major Errorcode   :- xecF_E_UNEXPECTED_RC
  Minor Errorcode   :- OK
  Probe Description :- AMQ6118: An internal WebSphere MQ error has
  occurred
  --------------------------------------------------------+
  MQM Function Stack
  zifMQCONN
  zifConnectToAgent
  zxcCreateECResources
  zutReadQMIniFile
  xcsBrowseIniCallback
  zutQMIniCallback
  zfpApiExitParser
  zfcApiExitReadCB
  xcsFFST
  
  On the sender channel side you may see the error "AMQ9503:
  Channel negotiation failed."
  
  This problem can occur in the following mq configuration: using
  threaded MCAs running multiple channels, with API exits
  configured, and where the threads are re-used.
  

Local fix

• None.  A code change will be incorporated in FixPack 5.3.1.9.
  

Problem summary

• Threaded applications using MQCNO_FASTPATH_BINDING connect
  option, were not connecting to SessionAnchor memset (except for
  the first thread). This causes the threads to fail if the queue
  manager has APIExits configured and the threads  use the
  SessionAnchor memset to load the APIExits. The problem will
  occur when the first thread in the application exits, and any
  subsequent thread connect and try to use the sessionAnchor to
  load APIExits.
  

Problem conclusion

• Modified code during thread initialization to connect to
  SessionAnchor memset for fastpath bound applications.
  

Temporary fix

• Ensure the first thread to connect in an application that uses
  MQCNO_FASTPATH_BINDING is the last to disconnect prior to the
  application ending.
  

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

  WEBS MQ NSS MIP

• Fixed component ID

  5724A3901

Applicable component levels

• R531 PSY

     UP