PH09848: RBA ON WRITE-ADD-COMPLETE AND READ-FOR-UPDATE DIFFER DURING DELETE PROCESSING ACROSS CI'S

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• This issue was discovered using the Infosphere Classic CDC
  product as it tries to do replication. A CICS task does an ESDS
  insert on a file. The task rolled back the UOW. After this we
  noticed an inequality in RBA's. The RBA on the
  WRITE-ADD-COMPLETE record would have one value, however, the
  RBA on the read-for-update for the logical delete processing is
  a different value. Classic CDC reads the ESDS update written in
  backout, which was logged with the RBA from the
  WRITE-ADD-COMPLETE and tries to locate the read-for-update, but
  since the RBA on the read-for-update is different the matching
  read-for-update cannot be located and replication stops.
  
  Additional Symptom(s) Search Keyword(s): KIXREVRJL
  

Local fix

• N/A
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All CICS Users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: During backout of a Write request the   *
  *                      corresponding read-update and           *
  *                      write-update log entries do not provide *
  *                      the same RBA for a non-RLS ESDS file.   *
  *                      This could affect products which        *
  *                      process replication log data such as    *
  *                      the Infosphere Classic CDC product.     *
  ****************************************************************
  LOGREPLICATE with LOGSTREAMID is specified against a non-RLS
  ESDS file. A write request is made against the file however the
  record's length means it will cross a CI boundary so the record
  is written at the start of the next CI. A SYNCPOINT ROLLBACK is
  issued so this write request is backed out. A read-update and
  write-update log entry is written due to the backout request.
  When using the replication log to recreate the system a
  write-update record is found but the corresponding read-update
  record cannot be found as the RBA being held in the read-update
  record is the predicted one rather then the new RBA that places
  the record in the new CI.
  
  Additional keywords: keysave keysave8 extended
  

Problem conclusion

• DFHFCVS1 has been updated so that a read-update log entry record
  contains the known RBA for a non-RLS ESDS file request when log
  replication is active against the data set and backout is in
  progress.
  

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:

  UI64624 UI64625 UI64626

Modules/Macros

• DFHFCVS
  

Fix information

• Fixed component name

  CICS TS Z/OS V5

• Fixed component ID

  5655Y0400

Applicable component levels

• R000 PSY UI64626

     UP19/08/08 P F908

• R100 PSY UI64625

     UP19/08/08 P F908

• R200 PSY UI64624

     UP19/08/08 P F908

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.