PK81405: Improved handling of larger resource files

APAR status

• Closed as program error.

Error description

• Use of the JAVA heap to load and store and process resources may
  over time cause heap fragmentation and potentially out of memory
  errors. Fragmentation typically occurs when authoring/uploading
  relatively large resources, and when syndicating those resources
  to another environment.
  

Local fix

• Avoid loading resources into memory and utilize streaming of
  resources when appropriate.
  

Problem summary

• Improved handling of larger resource files.
  Use of the java heap to load and store and process resources may
  over time cause heap fragmentation and potentially out of memory
  errors. Fragmentation typically occurs when authoring/uploading
  relatively large resources, and when syndicating those resources
  to another environment.
  

Problem conclusion

• Avoid loading resources into memory and utilize streaming of
  resources when appropriate.
  Must do / Recommended Tuning Configurations.
  
  1) If your doing some large files i.e. > 100MB per resource you
  may need to increase the Web containers transaction timeout from
  the default of 120 seconds.
  You can configure this on the WAS Admin console under
  Application Servers -> WebSphere_Portal -> Transaction Service
  -> Total transaction lifetime timeout.
  
  2) For unix environment, make sure the ulimit for file sizes
  isnt bound. Ensure the system has enough space to handle the
  larger disk requirements.
  
  3)  On DB2, make sure the database property
  fullyMaterializeLobData is set to false. If this isn't set, you
  will get Out of memory error's due to DB2 loading large
  resources into memory.
  
  4) For WCM, on WCMConfigService.properties, reduce the  the
  resourceserver.maxCacheObjectSize property from the default of
  10MB to 300kb or less.
  
  IMPORTANT NOTE:
  
  5) If you are seeing alot hung of web container threads from
  Portal's policy cache manager, then set:
  cacheinstance.com.ibm.wps.policy.services.PolicyCacheManager.lif
  etime= -1
  You can do this using the WAS Admin Console (Resource Provider
  ).
  
  This will reduce the db connections and load times and help
  alleviate hung threads.
  
  6) Enable transaction Syncronization by putting in
  "transaction.sync.remove=true" WCMConfigService.properties.
  
  7) Increase the Max Web Container threads to 100.
  
  8) The resource.maxUploadSize controls the max size of the
  uploaded files, recommended to be 100 MB or less.
  

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

  LOTUS WEB CONT

• Fixed component ID

  5724I2900

Applicable component levels

• R60H PSY

     UP