PI05212: Clients with short client retry timeouts experience TargetNotAvailable exceptions during server restart.

APAR status

• Closed as program error.

Error description

• When servers restart after it stops or fails over, clients with
  short retry settings (for example, 5 seconds or less) can
  experience TargetNotAvailable exceptions instead of routing
  quickly to the primary shards that are balanced to the
  restarted servers.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:  WebSphere eXtreme Scale users running       *
  *                  with short client retry time experience     *
  *                  TargetNotAvailable exceptions during server *
  *                  restart.                                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: Clients with short client retry         *
  *                      timeouts experience exceptions when     *
  *                      servers are restarted and shards        *
  *                      balance to new servers.                 *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When shards move to newly started servers, some existing
  clients do not get the updated route entries before the short
  retry timeouts (example, the client property
  requestRetryTimeout is set to 5 seconds). The clients
  experience unexpected TargetNotAvailable exceptions.
  The TargetNotAvailable exceptions indicate that the client
  could not find the correct endpoint; for example:
  BaseMap W CWOBJ0006W: An exception occurred:
  com.ibm.websphere.objectgrid.ObjectGridRuntimeException:
  com.ibm.websphere.objectgrid.TargetNotAvailableException:
  Retrying the request expired, the current timeout is set to
  4000 (ms). domain1:Grid:mapSet:591 forward: true
  

Problem conclusion

• Apply the fix. To work around this exception, extend the
  requestRetryTimeout property to a longer timeout. The
  TargetNotAvailable exceptions resolve after a short period of
  time.
  

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

• R850 PSY

     UP

• R860 PSY

     UP