PI83323: CICS SHUTDOWN HANGS AFTER MESSAGE DFHRM0131 RESYNCHRONIZATION REQUIRED WITH IRC RESOURCES.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• CICS shutdown hangs and region must be cancelled. Subsequent
  restarts all fail due to SOS condition, DFHSM0133. CICS must be
  started Initial to recover.
  .
  The last CICS shutdown message indicated that CICS was quiescing
  the RM domain:
  .
  DFHRM0131 RESYNCHRONIZATION REQUIRED WITH IRC RESOURCES
  .
  The next expected messages for the RM domain would be these two
  which were never issued:
  .
  DFHRM0204  There are no indoubt, commit-failed or backout-failed
    UOWs.
  DFHRM0130  Recovery manager has successfully quiesced.
  .
  Domain manager confirms that most domains are waiting for the RM
  domain to quiesce. The RM shutdown task is a CSXM task that is
  suspending/waking up every 10 seconds out of module DFHRMSL7,
  proc RMSL7_SLEEP_A_WHILE. Flag KEYPOINT_SCHEDULED is set to YES,
  so DFHRMSL7 believes that an activity keypoint is still in
  progress, and so waits for it to complete before doing a warm
  keypoint.
  .
  The joblogs showed that a CSKP task abended. After CKSP abended,
  there are no more messages indicating successful keypoints
  were taken, or that the log tail was trimmed. Expected messages
  would have been:
  .
  DFHRM0205 AN ACTIVITY KEYPOINT HAS BEEN SUCCESSFULLY TAKEN.
  DFHLG0743 TAIL OF LOG STREAM xxxxxx.DFHLOG DELETED AT BLOCK ID..
  .
  After CSKP abended, the KEYPOINT_SCHEDULED flag is left turned
  on, but CSKP is never resarted, no further keypoints are taken,
  and so the system log gets very large resulting in the
  subsequent SOS at startup. At startup, the SM domain shows that
  most of the storage is allocated to ECDSA as RMOFRMLK 's.
  

Local fix

• Correct the cause of the CSKP task abending.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS users                               *
  ****************************************************************
  * PROBLEM DESCRIPTION: An abend in a start of task TRUE for    *
  *                      the CSKP activity keypoint task         *
  *                      prevents subsequent keypoint            *
  *                      operations.                             *
  ****************************************************************
  CSKP is attached to perform an activity keypoint. A start of
  task TRUE is driven and this abends. CSKP terminates as a
  result. There is no mechanism to turn off the keypoint_scheduled
  flag in the RMSL anchor block. This prevents subsequent activity
  keypoints (or warm keypoint) from taking place. An emergency
  restart of the system can take an excessive time to complete as
  CICS has to read back a long way down DFHLOG to reconstruct the
  UOW activity in the previous run of the system. This can also
  result in a short on storage condition due to RMLK control
  blocks acquired during the extended restart.
  KEYWORDS: AKP msgdfhrm0001 dfhrm0001 rm0001 rm 0001 e/r
            system log class AKPs CSKPs
  

Problem conclusion

• CICS recovery manager has been changed to record the transaction
  token information for CSKP into the RMSL anchor block when the
  CSKP task is attached. This allows the keypoint indicator
  information to be cleared by backout processing if CSKP suffers
  an abend.
    The CICS TS 5.4 Supplementary Data Areas manual will be
  changed for the RMSL anchor as follows. The 40d bytes of
  reserved space at offset 55x will be modified to contain:
  "55    CHARACTER     3     *
         Prot
   58    CHARACTER     8     Keypoint_tran_token     CSKP task
   60    CHARACTER    29     *
         Prot"
  

Temporary fix

Comments

APAR Information

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

  PI78304

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

  UI48991

Modules/Macros

• DFHR2TRI DFHRMCD  DFHRMCD1 DFHRMCD2 DFHRMCI2 DFHRMCI3 DFHRMCI4
  DFHRMDM  DFHRMDU0 DFHRMDU2 DFHRMDU3 DFHRMDU4 DFHRMDU5 DFHRML1D
  DFHRMLK1 DFHRMLK2 DFHRMLK3 DFHRMLK4 DFHRMLK5 DFHRMLKQ DFHRMLN
  DFHRMLSD DFHRMLSF DFHRMLSO DFHRMLSP DFHRMLSS DFHRMLSU DFHRMNM
  DFHRMNM1 DFHRMNS1 DFHRMNS2 DFHRMOT  DFHRMR1D DFHRMR1E DFHRMR1K
  DFHRMR1S DFHRMRO  DFHRMRO1 DFHRMRO2 DFHRMRO3 DFHRMRO4 DFHRMROO
  DFHRMROS DFHRMROU DFHRMROV DFHRMRS  DFHRMSL  DFHRMSL1 DFHRMSL2
  DFHRMSL3 DFHRMSL4 DFHRMSL5 DFHRMSL6 DFHRMSL7 DFHRMSLF DFHRMSLJ
  DFHRMSLL DFHRMSLO DFHRMSLV DFHRMSLW DFHRMST  DFHRMST1 DFHRMTRI
  DFHRMU1C DFHRMU1D DFHRMU1E DFHRMU1F DFHRMU1G DFHRMU1J DFHRMU1K
  DFHRMU1L DFHRMU1N DFHRMU1Q DFHRMU1R DFHRMU1S DFHRMU1U DFHRMU1V
  DFHRMU1W DFHRMUC  DFHRMUO  DFHRMUW  DFHRMUW0 DFHRMUW1 DFHRMUW2
  DFHRMUW3 DFHRMUWB DFHRMUWE DFHRMUWF DFHRMUWH DFHRMUWJ DFHRMUWL
  DFHRMUWN DFHRMUWP DFHRMUWQ DFHRMUWS DFHRMUWU DFHRMUWV DFHRMUWW
  DFHRMXN2 DFHRMXN4 DFHRMXN5 DFHRMXNE
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R100 PSY UI48991

     UP17/07/22 P F707

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.