IC66155: BUILDING A HUGE INDEX USING INDEX PAGE LOGGING MAY RESULT IN SECONDARY SERVER TO PING TIMEOUT

APAR status

• Closed as program error.

Error description

• In a HDR environment using Index Page Logging Feature, building
  a huge index may result in "Ping Timeout". This is true because
  secondary is spending a large amout of time clearing pages,
  resulting in the logical log buffer to run full.
  
  The server will try to re-connect only after the clear pages are
  done on the secondary, which could be many hours after the "ping
  timeout" happened.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * Users using HDR and Index Page Logging                       *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * Building a huge index using index page logging may result in *
  * secondary server to ping timeout                             *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * Upgrade to 11.50.xC7 that contains the improvement.          *
  ****************************************************************
  

Problem conclusion

• The problem is caused by the secondary server been to busy to
  answer the ping messages from the primary. This is can happens
  more frequently if a huge index is built and Index Page Logging
  is used, in this particular case we identified an improvement
  when we can avoid to physical log new pages added to the
  partition which will reduce the load in the secondary server.
  

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

  IBM IDS ENTRP E

• Fixed component ID

  5724L2304

Applicable component levels

• RB15 PSN

     UP