PH55985: IMPORTED WORKLOAD IN CMAS STUCK IN FROZEN STATE

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• The workload was marked Frozen in the CMAS.  The job log
  of this CMAS shows that it was imported from another CMAS as
  a frozen workload:
  date time EYUWM0412I applid Workload (workload) from context
  (plexname) has been Frozen while importing from CMAS1.
  
  In the other CMAS joblog we see message confirming it was
  frozen:
  date time EYUWM0412I applid Workload (workload) from context
  (plexname) has been Frozen due to loss of contact with CMAS2.
  
  This EYUWM0412I message is written when either a CMAS or MAS has
  terminated while the workload was active on either type of
  region, or a CMAS that is being joined to an existing workload
  has recognized that the workload is currently in a frozen state.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICSPlex SM Users.                       *
  ****************************************************************
  * PROBLEM DESCRIPTION: Workloads may incorrectly show as       *
  *                      frozen on CMAS regions participating    *
  *                      in a workload after the down regions    *
  *                      have been restored.                     *
  ****************************************************************
  * RECOMMENDATION: After applying the PTF that resolves this    *
  *                 APAR, all CMAS regions must be restarted.    *
  *                 These restarts may occur in any order. .     *
  ****************************************************************
  When a CMAS which has connected MAS regions participating in a
  workload is stopped, these regions are marked as lostcon (in
  the associated WAOR/WTORs), and the workload is frozen. If a
  another CMAS has its first target region in a workload start, it
  will attempt to import the existing workload from another active
  CMAS participating. If the down CMAS restarts during this import
  process (after the other CMAS has exported its workload state),
  the status of affected MAS in the imported workload information
  may be out of sync with the local Topology data held in the
  importing CMAS.
  
  This causes the workload freeze count(WRKD_FREEZE_CNT) to
  include the previous lostcon MAS regions. When the WAOR and
  WTOR values are being imported from the WRKQ, the import process
  in EYU0WMQU, due to the previous lostcon state, the local WTOR
  and WAOR values have the lostcon flag set, and then attempts
  to update topology for these regions. This calls TSNC to update
  topology, which fails. This results in a mismatch between the
  workload and topology for these regions. As such, the workload
  remains frozen incorrectly, due to the count not being able to
  be reduced for these AOR/TORs.
  
  This results in a mismatch between CMAS regions in the workload
  being frozen.
  

Problem conclusion

• Workload import processing in EYU0WMQU has been corrected, to
  reconcile the incoming workload state for the
  WAOR/WTORs with the local topology state.
  This correctly reconciles the freeze count (WRKD_FREEZE_CNT)
  for active/lostcon MASes during import, and correctly
  freezes/defrosts the workload during import.
  

Temporary fix

Comments

APAR Information

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

  PH52540

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

  UI93035 UI93036

Modules/Macros

• EYU0WMQU
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R20M PSY UI93036

     UP23/08/05 P F308

• R30M PSY UI93035

     UP23/08/05 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.