PH25026: DFHAP0001 DBDCCICS ABEND (CODE ---/AKEA) AT OFFSET X'1DD6' IN DFHD2EX1. CAN HAPPEN WHEN SHUTTING DOWN DB2.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• You see the following message:
  .
  DFHAP0001 DBDCCICS An abend
  (code ---/AKEA) has occurred at offset X'1DD6' in module
  DFHD2EX1.
  .
  Analysis of the resulting dump shows that the abend
  AKEA is caused by an abend0C4 in DFHD2D2 because Reg8 is bad.
  The offset into DFHD2D2 of the 0C4 can vary. One common offset
  is x'402'
  .
  This problem can happen when there is an abend of
  an L8 TCB while it is being used to process an SQL request.
  This typically happens when DB2 is being forced down. For
  example:
  .
  -STOP DB2 MODE (FORCE)
  .
  It can also happen when you
  are shutting down the DB2CONN with:
  .
  CEMT SET DB2CONN
  NOTCONNECTED FORCE
  .
  It can also happen when you Purge a task
  while it is in DB2 processing an SQL request.
  .
  Additional
  Symptoms:
  .
  The *REC CSB trace is missing the '*REC'. Instead
  you see x'000A0000'.
  .
  Additional keywords:
  R8 CSUB D2D2REC
  00000402
  

Local fix

• Turn on CICS internal trace and ensure that the RI trace
  component it set to trace level 1.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS Users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: DFHAP0001 Abend AKEA at offset x'1DD6'  *
  *                      in DFHD2EX1 .                           *
  ****************************************************************
  Following the purge of a CICS-DB2 transaction, an abend 0C4 is
  seen in DFHD2D2's recovery routine. This is then followed by
  the AKEA abend in DFHD2EX1.
  The problem is being caused when the kernel sets up to drive
  DFHD2D2's recovery routine; it is calling DFHKERRI in AMODE 31
  and not AMODE 64. This then results in not all the base
  registers for DFHD2D2 being setup correctly.The problem only
  presents itself if DFHD2D2 does not issue a trace call when
  entered, as doing so causes the registers to be saved at that
  point.
  Keywords: D2D2 F4SA set_recovery 80000000 00000001
  

Problem conclusion

• The kernel domain has been changed to always call DFHKERRI in
  AMODE 64.
  

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:

  UI70733

Modules/Macros

• DFHDSIT  DFHKERN
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R200 PSY UI70733

     UP20/07/27 P F007

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.