PM92210: A CLIENT REQUEST ENCOUNTERS A TARGETNOTAVAILABLEEXCEPTIONWHEN US ING FAIL FAST AND NOT RETRYING AFTER ROUTING ERROR.

APAR status

• Closed as program error.

Error description

• A client request encounters a TargetNotAvailableException with
  a caused by exception of org.omg.CORBA.NO_RESPONSE or
  MessageTimeOutException when using fail fast.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:  WebSphere eXtreme Scale users that are      *
  *                  running                                     *
  *                  in fail fast mode. The clients do           *
  *                  not retry after a routing failure.          *
  ****************************************************************
  * PROBLEM DESCRIPTION: After an initial                        *
  *                      TargetNotAvailableException exception,  *
  *                      the                                     *
  *                      client cannot route requests for a      *
  *                      period of time and issues               *
  *                      TargetNotAvailableException exception.  *
  *                      During                                  *
  *                      normal recovery, the target that        *
  *                      failed with                             *
  *                      TargetNotAvailableException is          *
  *                      removed on the client in anticipation   *
  *                      of acquiring updated routing            *
  *                      information. In a fail-fast scenario    *
  *                      with NORESPONSE, the client did not     *
  *                      get an updated route for a period of    *
  *                      time.                                   *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  The client runs fail-fast mode when the client-side
  requestRetryTimeout  property  is set to 0.
  A client request encounters a
  TargetNotAvailableException exception, caused by a
  org.omg.CORBA.NO_RESPONSE or MessageTimeOutException exception.
  After
  one of these exceptions occurs, the client continues to
  experience a TargetNotAvailableException exception with the
  message
  explanation of "Cannot find primary for
  gridName:mapSetName:partitionID."
  Example:
  com.ibm.websphere.objectgrid.TargetNotAvailableException:
  Cannot find primary for DefaultDomain:Grid:mapSet:10.
  However, the partition reported in the
  TargetNotAvailableException exception is still running and
  reports as  available in the xscmd -c routetable command.
  

Problem conclusion

• Use one of the following solutions to work around the errors:
  * For the client to recover naturally, issue an xscmd -c
  triggerPlacement command. This recovery can take up to
  10 minutes.
  * Disconnect and reconnect the
  client to refresh the client side routing information. Apply the
  fix.
  

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

  WS EXTREME SCAL

• Fixed component ID

  5724X6702

Applicable component levels

• R860 PSY

     UP