JR45722: INSTANCES WIDGET CAUSES OUT OF MEMORY CONDITION IN REST SERVICES WITH LARGE NUMBER OF INSTANCES

IBM Business Monitor V8.0.1 Fix Pack 2 (8.0.1.2)
IBM Business Monitor V7.5.1 Fix Pack 2 (7.5.1.2)
Version 8.5 Refresh Pack 5 for the IBM Business Process Manager products

APAR status

• Closed as program error.

Error description

• Retrieval of instance data by the Instances widget (from
  Business Monitor REST services) triggers a memory leak in REST
  services.  Over time this can result in an out of memory
  condition in the Java Virtual Machine (stand-alone server or
  cluster) where REST services is deployed, especially with a
  large number of instances.
  

Local fix

• N/A
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:  Users of WebSphere Business Monitor         *
  *                  Version 7.0, and IBM Business Monitor       *
  *                  Version 7.5 and 8.0.                        *
  ****************************************************************
  * PROBLEM DESCRIPTION: Retrieval of instance data by the       *
  *                      Instances widget (from Business         *
  *                      Monitor REST services) triggers a       *
  *                      memory leak in REST services.  Over     *
  *                      time this can result in an out of       *
  *                      memory condition in the Java Virtual    *
  *                      Machine (stand-alone server or          *
  *                      cluster) where REST services is         *
  *                      deployed, especially with a large       *
  *                      number of instances.                    *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  The pagination method used by REST services used a very
  inefficient method to go page by page through the instances.
  In a high volume environment, this caused an out of memory
  condition due to the size of the recordsets being opened and
  traversed.
  

Problem conclusion

• Improving the query pagination method for each supported
  database type (except Derby) resolved the issue.  (For Derby,
  improvement is not possible; however, this should not be an
  issue as Derby is not recommended for a high volume production
  environment.)
  
  An interim fix for WebSphere Business Monitor Version 7.0.0.4
  has been published to Fix Central.  Please refer to the
  separate file that is automatically downloaded with the
  interim fix for prerequisite information and
  installation/uninstallation instructions.  For IBM Business
  Monitor Version 7.5 and Version 8.0, the issue will be
  resolved in future maintenance releases.
  

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

  WEB BUS MONITOR

• Fixed component ID

  5724M2400

Applicable component levels

• R700 PSY

     UP

• R750 PSY

     UP