PI38797: JOB PURGE ISSUE FOR WSGRID (JAVA) JOBS AND GETTING SERVANT ABEND

APAR status

• Closed as program error.

Error description

• From time to time, customer was getting a servant ABEND=SEC3
  U0000 REASON=04130007.
  
  Following exception was logged into the control region:
  
  BBOO0327I HTTP REQUEST
  TIMEOUT:(0329):(FF595963):(0001001D):(009BC1E0):
  (S0778880):(2015/02/11 15:04:39.374135):(2015/02/11
  15:04:39.374328):
  (2015/02/11 15:05:41.535976):(RemoteWebContainer):(httprequest):
  (ipaddr=<ip address> port=21575):():():
  (/PGCController/PGCControllerServlet)
  BBOO0232W A request for Class Name 'RemoteWebContainer' and
  Method Name
  'httprequest', from ip addr=<ip address> port=21575, has timed
  out.
  The servant process associated with the request will be
  terminated.
  Request Id(FF595963)
  BBOO0303I ADDITIONAL INFORMATION FOR HTTP REQUEST
  TIMEOUT: /PGCController/PGCControllerServlet
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:  All users of WebSphere Compute Grid         *
  *                  version 8.0.0.4.                            *
  ****************************************************************
  * PROBLEM DESCRIPTION: An HTTP request timeout occurs during   *
  *                      startup of a servant region. The HTTP   *
  *                      timeout is associated with              *
  *                      initialization of the                   *
  *                      PGCControllerServlet.                   *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When running with runIOImprovementForWSGrid enabled, the
  JobSchedulerMDB requests job logs for jobs that have been
  dispatched to the endpoint, but are not yet executing.  This
  is resulting in all internal worker threads being tied up
  waiting for the PGCControllerServlet initialization to complete
  causing a deadlock situation. At the same time, an
  HTTP request is coming into the servant which is waiting for
  the same lock to free up. This request encounters a timeout
  

Problem conclusion

• An update was implemented in the JobSchedulerMDB
  runIOImprovementForWSGrid code path to ensure that job logs
  are not fetched for jobs in dispatched state.
  
  The fix for this APAR is currently targeted for inclusion in
  fixpack 8.0.0.5.  Please refer to the Recommended Updates page
  for delivery information:
  http://www.ibm.com/support/docview.wss?uid=swg27022998
  

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

  WXD Z COMP GRID

• Fixed component ID

  5655V6201

Applicable component levels

• R800 PSY

     UP