PH30083: DB2 ABNORMAL TERMINATION MAY RESULT FOLLOWING ABND=04E-00E50001

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• DDF thread abend with reason code
  00E50001 may cause Db2 abnormal
  termination with reason codes
  00E50705 or 00E50727.
  
  DSNL033I may also be seen reporting
  ABEND 0C4 REASON 00E50001
  
  The following offsets are V12
  04E-00E50001 DSNSFBK    +2364
  04E-00E50001 DSNSRSUP  +01A7A
  
  
  
  NOTE: This applies to Db2 12 only
  
  Additional keywords and symptoms:
  ********************************
  DB2DDF DSNSFBK DSNSRSUP
  00E50705 RC00E50705
  00E50727 RC00E50727
  00E50001 RC00E50001
  DSNL033I DSNV086E LC32
  ABTERM DB2TERM
  OFFSET2364 OFFSET01A7A
  RESYNC01 028.RESYNC01
  

Local fix

• no local workaround or fix
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * All Distributed Data Facility (DDF) users.                   *
  * Especially those who access Db2 via TCP/IP.                  *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * Abend 04E-00E50001 DSNSFBK+2364                              *
  * or Abend 04E-00E50001 DSNSRSUP+01A7A                         *
  * with possible subsequent Db2 abnormal                        *
  * termination with reason codes                                *
  * 00E50705 or 00E50727.                                        *
  ****************************************************************
  * RECOMMENDATION:                                              *
  * Apply corrective PTF when available                          *
  ****************************************************************
  Distributed application environments have the ability, whether
  intentional or due to environmental reasons, to cancel their
  last request to Db2 for z/OS by canceling the TCP/IP
  connection with Db2 for z/OS. When Db2 detects such a client
  connection cancel request, a service task within the ssidDIST
  address space is dispatched to handle the cancel of the
  current request being processed. In most situations where Db2
  has not exhausted its pool of available storage blocks to
  control synchronization between tasks, the cancel process
  completes normally. However, in rare situations, such as this
  reported case, where Db2 has currently exhausted its pool of
  available control blocks, to complete the cancel of the
  current request being processed, Db2 must first acquire more
  blocks. The allocation of more storage for the additional
  blocks must be done under a latch for the benefit of
  serialization. The latch to acquire storage was of the same
  latch class as a latch already held by the task attempting to
  handle the client connection loss request. This situation then
  leads to ABEND 04E-00E50001 at DSNSFBK+2364 or DSNSRSUP+01A7A
  which may then lead to a subsequent Db2 abnormal termination
  with reason codes 00E50705 or 00E50727.
  

Problem conclusion

• Db2 for z/OS client connection cancel processing has been
  changed to use a latch which will not conflict with the latch
  needed for the allocation of new storage blocks to be used for
  synchronization between tasks within Db2.
  

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:

  UI73504

Modules/Macros

• DSNLILNR DSNLILOS DSNLQINA
  

Fix information

• Fixed component name

  DB2 OS/390 & Z/

• Fixed component ID

  5740XYR00

Applicable component levels

• RC10 PSY UI73504

     UP21/01/23 P F101 ¢

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.