PH06344: IMPROVE DDF THREAD TERMINATION PROCESSING WHEN A CONNECTION IS ABNORMALLY CLOSED.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Improve DDF thread termination processing when a connection is
  abnormally closed.
  See PH06075 for Db2 11 for z/OS.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * All Distributed Data Facility (DDF) users                    *
  * of Db2 for zOS.                                              *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * Db2/DDF is unable to service work if                         *
  * remote client environments continually                       *
  * create and terminate connections to                          *
  * the Db2 z/OS server.  This occurs                            *
  * because Db2/DDF authentication related                       *
  * processing is unable to keep up with                         *
  * the abundance of failed connection                           *
  * events.                                                      *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * Apply corrective PTF when available                          *
  ****************************************************************
  A large number of DDF connection requests were queued up for
  one of the authorization service tasks, and were canceled
  before they were selected to be processed.  The cancels
  resulted in an ABEND04E RC00E50013 under the authorization
  service task.  The cancel abend was passed back to the
  requesting DDF agent, which also abended with ABEND04E
  RC00E50013.  DDF processing successfully cleaned up the DDF
  agent.
  The authorization service task abends, and subsequent abend
  recovery and service task reinstatement, was driving up TCB
  time in the DB2-DIST address space.  This process was occurring
  under all the authorization service task TCBs concurrently,
  resulting in contention which elongated the time that it took
  to recover from the canceled requests.
  

Problem conclusion

• Service task cancel processing has been improved to avoid
  abending when a queued request has been canceled prior to being
  selected for processing.  Instead of abending, and allowing the
  service task TCB to terminate, the service task controller will
  pass the cancel back to the service task requestor and will
  continue processing with the next request on the service task
  queue.  See PH06075 for Db2 V11 fix.
  

Temporary fix

• *********
  * HIPER *
  *********
  

Comments

APAR Information

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

  PH06075

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

  UI60431

Modules/Macros

• DSNVEUS3
  

Fix information

• Fixed component name

  DB2 OS/390 & Z/

• Fixed component ID

  5740XYR00

Applicable component levels

• RC10 PSY UI60431

     UP19/01/08 P F901 ¢

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.