PI45204: POTENTIAL PERFORMANCE ISSUE ON WAS ND CLUSTER ENVIRONMENT DURING THE RUNTIME SYNCHRONIZATION

IBM MobileFirst Platform Foundation Interim Fix README for 7.0
IBM MobileFirst Platform Foundation Interim Fix README for 6.3
IBM Worklight Interim Fix README for 6.2.0.1

APAR status

• Closed as program error.

Error description

• When WAS cluster members are started in parallel after one is
  already started, during the Worklight runtime startup the
  downloading of the artifacts for all the members  use the same
  HTTP endpoint (the one of the started member) instead of using
  their own endpoint so provoking a bottleneck in case there are a
  lot of artifacts to download and several members and WAS server
  very long startup (more than 10 mns).
  

Local fix

• Stop/start the cluster in order that all members start in
  parallel.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * Worklight administrator users.                               *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * When WAS cluster members are started in parallel after one   *
  * is already started, the downloading of the artifacts during  *
  * the runtime synchronization use the same HTTP endpoint (the  *
  * one of the started member) for every cluster member instead  *
  * of using the own endpoint of the cluster member so provoking *
  * a bottleneck in case there are a lot of artifacts to         *
  * download and several members and increasing the WAS server   *
  * startup (can be more than 10 mns to achieve the runtime      *
  * synchronization).                                            *
  * The workaround is to stop/start the cluster in order that    *
  * all members start in parallel.                               *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * -                                                            *
  ****************************************************************
  

Problem conclusion

• When a cluster member start if the admin service is deployed in
  the same cluster than the runtime (symmetric deployment), the
  HTTP(s) endpoint to download the Worklight applications uses the
  port and the hostname of this cluster member.
  

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

  MFPF/WORKLIGHT

• Fixed component ID

  5725I4301

Applicable component levels

• R620 PSY

     UP

• R630 PSY

     UP

• R700 PSY

     UP