PH54493: DFHDS0001 0C4 AKEA IN DFHDSKE AFTER DB2 CRASHED

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Running CICS Transaction server 6.1, with transactions making
  DB2 calls, when DB2 goes down.
  In CICS , you see the following error:
  DFHDS0001  An abend (code 0C4/AKEA) has occurred at offset
             X'047E' in module DFHDSKE.
  The program check is  due to R1 containing an invalid address.
  
  R1 was initially loaded from a DFHDSDS_TCB control block,
  that represents an L8 TCB.
  The problem is, the TCB has been detached, so its no longer a
  valid TCB address, yet CICS still has
  control blocks pointing to it - like KTCB
  
  Here are some exception trace entries:
  DS 0220 DSDS4 *EXC* DSDS4_PURGE_EXCP1
                      PURGE,FORCE_CANCEL,4C71EB00,0042591
  DS 0220 DSDS4 *EXC* DSDS4_PURGE_EXCP1
                      PURGE,FORCE_CANCEL,4C720E00,0042042
  DS 0167 DSIT  *EXC* PULL_RUNNING_TASK_TO_QR
                      PROCESS_DEAD_TCBS,0044191
  DS 0167 DSIT  *EXC* PULL_RUNNING_TASK_TO_QR
                      PROCESS_DEAD_TCBS,0044192
  
  AP 31D1 D2STR *EXC* - DB2_NOT_ACTIVE
  DS 0049 DSKE  *EXC* - DSKE_RECOVERY - FUNCTION(TASK_REPLY)
                 TASK_TOKEN(2E642000) ATTACH_TOKEN(4C720E00)
  
  
  Additional symptoms:  KIXREVDAM
  detach open standby abend AKEX AKEH AEXZ ADZ3 04E
  freemained TCB part of DFE
  Additional keywords: abend0c4 dfhdske , abends0c4
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: CICS is pointing to an address of a TCB *
  *                      that no longer exists. The data in this *
  *                      location causes an 0C4.                 *
  ****************************************************************
  Tasks running on L8 or L9 TCBs are abending and their ESTAE is
  driven for each one.
  PROCESS_DEAD_TCBS runs for each task twice, causing the
  ESTAE WAIT ECB to be posted and the TCBs to complete
  termination before the associated tasks.
  Task termination checks the associated L8/L9 TCB but the TCB
  address within the DS_TCB is no longer valid, so the data
  found there can cause 0C4s.
  

Problem conclusion

• DFHDSKE has been changed to check that the TCB is still
  available before attempting to refer to its address.
  

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:

  UI93240

Modules/Macros

• DFHDSKE
  

Fix information

• Fixed component name

  CICS TS Z/OS V6

• Fixed component ID

  5655YA100

Applicable component levels

• R400 PSY UI93240

     UP23/08/22 P F308

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.