PI39731: WMQ Z/OS:LOOP OCCURRED WHEN STOPPING WAS SERVER AFTER ADMIN STRUCTURE FULL CONDITION.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• There is a loop problem when stopping WAS server
  after Admin Structure Full condition.
  The error in the recovery processing for the initial 5C6 abend
  can lead to an ACE being incorrectly left on the VLCA chain for
  the application address space. When the Servant region was
  subsequently  stopped, CSQMCTXE added the context ACE to the
  VLCA chain, however  because it was already on the chain, this
  led to the reported loop  condition.
  
  
  Additional Symptom(s) Search Keyword(s):
  

Local fix

• n/a
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of WebSphere MQ for z/OS Version 8 *
  *                 Release 0 Modification 0.                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: High CPU loop in MQ application address *
  *                      space during termination processing     *
  *                      following an earlier abend in RRS exit  *
  *                      processing for that address space.      *
  *                      The application address space hangs and *
  *                      cannot be canceled.                     *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  While an MQ RRS Exit is processing a RRS context (for example,
  to prepare or commit work done under the context) on behalf of
  an RRS application e.g. WSAS, the context ACE is added to the
  VLCA allied ACE chain for the address space.
  If the RRS exit abends, CSQ3RRSF is invoked to provide recovery
  processing, including removing the ACE from the VLCA chain,
  however an error in this recovery routine can lead to the ACE
  being incorrectly left on the chain.
  If the ACE is subsequently used again (for example during end
  context processing), it can be added to the VLCA chain again,
  leading to a loop being introduced to the chain.
  When the chain is subsequently processed, for example by end of
  task processing, or other RRS exits for the address space, the
  loop in the VLCA chain results in processing looping, causing
  high cpu.
  A latch is held while processing the VLCA chain, so any other
  tasks attempting to use this chain will be suspended due to
  the looping task holding this latch. This can result in the
  address space failing to end correctly, even when canceled.
  

Problem conclusion

• CSQ3RRSF is changed to correctly remove the context ACE from the
  VLCA chain when needed.
  000Y
  CSQ3RRSF
  

Temporary fix

• *********
  * HIPER *
  *********
  

Comments

APAR Information

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

  PI39495

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

  UI27544

Modules/Macros

• CSQ3RRSF
  

Fix information

• Fixed component name

  WMQ Z/OS 8

• Fixed component ID

  5655W9700

Applicable component levels

• R000 PSY UI27544

     UP15/06/03 P F506 ¢

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.