PH06247: MQ Z/OS: ABEND5C6 REASON 00D10231 IN CSQJW004 WHEN THE HEADER PAGE OF A PAGE SET HAS AN OUT-OF-RANGE RBA.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A dump occurred with title:
  ABN=5C6-00D10231,U=SYSOPR  ,C=MQ900.900.RLMC-CSQJCKLO,
    M=CSQJWE01,LOC=CSQJL002.CSQJW004+000005D2
  
  The queue manager ends with:
    CSQV086E QUEUE MANAGER ABNORMAL TERMINATION REASON=00E50702
  
  This abend can occur if at some time in the past the queue
  manager had its logs recreated without CSQUTIL RESETPAGE being
  run against each page set.
  
  This is similar to the situation in PI97758 (MQ V8)/ PI97758
  (MQ V9).  Those APARs detect when there are dataset consistency
  issues on spacemap pages on a page set.
  
  This APAR will detect a different type of issue with the header
  page having an excessively large RBA.
  
  
  L2 procedure to find the number of the page set having the
  problem:
  
  - If MSTR trace is active, look at the 3rd word in the DWP2MMRA
  DATA trace entry leading up to the abend.  That word contains
  the hex value of the page set, which will need to be converted
  to decimal.
  
  - If MSTR trace is not active:
    - IP ST FAILDATA.  Get Reg13.
    - Reg13+x'4' points to the prior Register Save Area (RSA) for
      CSQPDWP.
    - In that prior RSA, get the saved Reg1 from +x'18'.
    - Get the hex PSID number from that Reg1+x'10' and the RBA
      from the address pointed to by that Reg1+x'0'.
    - The RBA value will be higher than the RBAs listed in the
      MSTR joblog, for example the RBA in the CSQJ099I message.
  
  Additional Symptom(s) Search Keyword(s):
  CSQJW004 CSQJWE01
  

Local fix

• If the prior use of the page set was from a normal queue
  manager shutdown, CSQUTIL RESETPAGE will need to be performed
  for the affected page set. See https://ibm.biz/Bd2dpC
  .
  If the RESETPAGE command fails, then:
   - For page sets other than 0:
      Plan to stop the queue manager as soon as it is convenient,
      and follow the procedure to restore the page set(s) to a
      consistent state:
  
      - Run CSQUTIL with SCOPY PSID(x) to save persistent
        messages on the page set to a data set
      - Format the page set with TYPE(NEW)
      - Start the queue manager and reload the messages from
        the data set using SCOPY LOAD
   - For page set 0, contact IBM Service.
  .
  .
  To prevent this problem with the RBA, the correct steps for
  recovery are in topic "Restarting if you have lost your log
  data sets" at https://ibm.biz/BdZhsP
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of IBM MQ for z/OS Version 9       *
  *                 Release 0 Modification 0 and Release 1       *
  *                 Modification 0.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: Abend 5C6-00D10231 occurs during        *
  *                      startup after migration from an earlier *
  *                      release, if the queue manager           *
  *                      had previously been cold started, or    *
  *                      had its logs reset, and the procedure   *
  *                      was not fully completed.                *
  ****************************************************************
  During start up CSQIECUR runs to examine each available page set
  to extract the logRBA that Log processing must be started from
  to correctly rebuild the page set. In the reported problem the
  header page of one of these page sets contained an RBA that was
  higher than should be expected, but the problem was not
  identified immediately.
  
  When these updates were written out by the deferred write
  processor, it detected that the pages contained a log RBA higher
  than the current highest log RBA for the queue manager - this is
  an invalid situation and so the queue manager is terminated.
  
  The invalid RBA occurs if the queue manager had been cold
  started or had its logs reset, and the procedure was not
  completed for all page sets (for example, a page set was not
  formatted during a cold start, or RESETPAGE was not run against
  a page set during a log reset operation).
  
  Prior to PI69479 at V8 the incomplete operation was not detected
  and the queue manager was able to start, despite the invalid
  RBAs. It was then able to run normally (provided no recovery
  actions were required on the pages with invalid RBAs) with no
  indication of there being any problem.
  
  It is only when migrating to V9 that the invalid RBA is detected
  and the queue manager is unable to start.
  

Problem conclusion

• CSQIECUR is changed to detect header pages with an invalid
  RBA, and issue new message CSQI053E for each affected page set,
  allowing the appropriate recovery action to reintroduce the
  page set in a consistent state to be planned.
  
  In addition to issuing CSQI053E, affected page sets are
  suspended, allowing the queue manager to start. Queues on
  affected page sets will not be accessible until the page set is
  recovered.
  
  The knowledge centre is updated as follows:
  
  Data manager messages (CSQI...)
  https://www.ibm.com/support/knowledgecenter/SSFKSJ_9.0.0/
                                   com.ibm.mq.ref.doc/csq_i.htm
  https://www.ibm.com/support/knowledgecenter/SSFKSJ_9.1.0/
                                   com.ibm.mq.ref.doc/csq_i.htm
  
  New message CSQI053E is added:
  CSQI053E
   Invalid page RBA found during restart for page set <psid>
  
  Severity
      8
  Explanation
  
      A page containing an invalid RBA was detected on the
      indicated page set during startup, indicating the page set
      is not in a consistent state.
      This is normally as a result of the page set not being
      correctly processed during a past cold start operation or
      RESETPAGE operation.
  
  System action
  
      The page set is suspended. Queues using the page set will
      be inaccessible until the queue manager is started with the
      page set in a consistent state.
  
  System programmer response
  
     When <psid> specifies page set 0:
        Contact IBM Service
  
     For page sets other than 0:
      Plan to stop the queue manager as soon as it is convenient,
      and follow the procedure to restore the page set(s) to a
      consistent state:
  
      - Run CSQUTIL with SCOPY PSID(x) to save persistent
         messages on the page set to a data set
      - Format the page set with TYPE(NEW)
      - Start the queue manager and reload the messages from
         the data set using SCOPY LOAD
  

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:

  UI61973 UI61974 UI61975 UI61976 UI61977 UI61978 UI61979 UI61980
  UI61981 UI61982 UI61983 UI61984

Modules/Macros

• CSQFIDIC CSQFIDIE CSQFIDIF CSQFIDIK CSQFIDIU CSQFMTXC CSQFMTXE
  CSQFMTXF CSQFMTXK CSQFMTXU CSQIECUR
  

Fix information

• Fixed component name

  IBM MQ Z/OS V9

• Fixed component ID

  5655MQ900

Applicable component levels

• R000 PSY UI61973

     UP19/04/11 P F904 ¢

• R001 PSY UI61974

     UP19/04/12 P F904 ¢

• R002 PSY UI61975

     UP19/04/12 P F904 ¢

• R003 PSY UI61976

     UP19/04/12 P F904 ¢

• R004 PSY UI61977

     UP19/04/12 P F904 ¢

• R005 PSY UI61978

     UP19/04/12 P F904 ¢

• R100 PSY UI61979

     UP19/04/11 P F904 ¢

• R101 PSY UI61980

     UP19/04/11 P F904 ¢

• R102 PSY UI61981

     UP19/04/11 P F904 ¢

• R103 PSY UI61982

     UP19/04/11 P F904 ¢

• R104 PSY UI61983

     UP19/04/11 P F904 ¢

• R105 PSY UI61984

     UP19/04/11 P F904 ¢

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.