PM56180: Asynchronous replication can get behind and not catch up, while under continuous load.

APAR status

• Closed as program error.

Error description

• Asynchronous replication does not update enough data on each
  replicating cycle when grids contain a large number of maps.
  While under continuous load, the replica cannot stay current.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:  Users of WebSphere eXtreme Scale who have   *
  *                  several maps in their data grids and use    *
  *                  asynchronous replication.                   *
  ****************************************************************
  * PROBLEM DESCRIPTION: Asynchronous replication can fall       *
  *                      behind when the primary is under load.  *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When reviewing the revisions information for a primary and
  replica, the replica can fall behind, even though the primary
  is able to stay current with a large load.  When many maps
  exist, the replication code is conservative to prevent running
  out of memory when replicating.  Therefore, the replication
  packets sent to the replica can be smaller.
  

Problem conclusion

• The code was updated to make sure more data is sent during
  each replication cycle.
  

Temporary fix

Comments

APAR Information

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

  PM56174

• 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

• R711 PSY

     UP