PI76882: MQ Z/OS: THE QUEUE MANAGER SHOULD PROVIDE A WARNING IF THE ACTIVE LOG DATA SET IS LARGER THAN THE ALLOWED 4GB

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• The maximum active log data set size is 4 GB as documented at
  https://www.ibm.com/support/knowledgecenter/SSFKSJ_8.0.0/com.ibm
  .mq.pla.doc/q006060_.htm
  
  If the allocation for the active log is greater than 4GB, MQ
  does not report any errors and only uses the log size that
  exceeds 4GB.  For instance, if you allocate 4136MB for each
  Active Log, the difference in the START and END RBA values in
  the MSTR log messages or CSQJU004 output indicate MQ is just
  using 40MB (4136 - 4096).
  
  This will result in faster than expected active log switching:
   CSQJ002I END OF ACTIVE LOG DATA SET ...
   CSQJ001I CURRENT COPY 1 ACTIVE LOG DATA SET IS ...
  
  It may result in long-running units of work being detected more
  quickly:
   CSQJ160I CSQJLRUR LONG-RUNNING UOW FOUND,
     URID=<urid> CONNECTION NAME=<connection id>
   CSQR026I Long-running UOW shunted to
     RBA=<rba>, URID=<urid> connection name=<connection id>
  
  Archiving using OFFLOAD=YES in CSQ6LOGP is recommended, but if
  it is NO (off), logs may switch too quickly for log shunting to
  keep up, resulting in
    CSQR027I Long-running UOW shunting failed,
    URID=<urid> connection name=<connection id>
  and if a backout / rollback is then attempted:
    CSQJ113E RBA 0000003F06178000 NOT IN ANY ACTIVE
    OR ARCHIVE LOG DATA SET, CONNECTION-ID=C1A4CHIN
  
    CSQV086E QUEUE MANAGER ABNORMAL TERMINATION REASON=00D94001
  
    DUMP TITLE=ssid,ABN=5C6-00D1032A,U=MQUSER  ,C=W9700.800.RLMC-
    CSQJLGR ,M=CSQJRE01,LOC=CSQJL002.CSQJR003+00000EEA
  
  
  Note that fast log switches may be due to a high-volume or
  looping application. The long-running UOW symptoms indicate an
  application that is not committing frequently enough.  The
  CSQR027I may also result from having too few active logs
  defined.
  
  
  Additional Symptom(s) Search Keyword(s):
  ABEND5C6 ABENDS5C6 S5C6 S05C6 5C6
  ABEND0C4 ABENDS0C4 S0C4 S00C4 0C4 3B in CSQFPMS2
  

Local fix

• As stated at
  https://www.ibm.com/support/knowledgecenter/SSFKSJ_8.0.0/com.ibm
  .mq.pla.doc/q006060_.htm
  
  When using striped data sets, where the data set is spread
  across multiple volumes, the specified size value is allocated
  on each DASD volume used for striping. So, if you want to use 4
  GB logs and four volumes for striping, you should specify:
  
      CYLinders(1456)
      Megabytes(1023)
  

Problem summary

• ****************************************************************
  * USERS AFFECTED:                                              *
  * All users of IBM MQ for z/OS Version 9 Release 0             *
  * Modification 0.                                              *
  ****************************************************************
  * PROBLEM DESCRIPTION:                                         *
  * Problem 1:                                                   *
  * If active LOG datasets are allocated using Extended VSAM     *
  * feature and the allocated space exceeds 4GB the RBA range    *
  * available is smaller than the allocated space. This can      *
  * cause an increased checkpoint frequency.                     *
  *                                                              *
  * ABEND 5C6-00D1032A in CSQJR003 can be generated when running *
  * without archive logging as result of running low in space in *
  * the active logs. In this case the manager terminated with    *
  * reason 00D94001.                                             *
  *                                                              *
  * Problem 2:                                                   *
  * ABEND 0C4-3B in CSQFPMS2.                                    *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  Problem 1:
  When a queue manager opens an active log dataset that has been
  allocated with space size larger than 4GB, the queue manager
  receives a truncated 32 bit value of the highest RBA allocated
  resulting in the smaller RBA range used by the queue manager.
  
  Problem 2:
  When CSQJW207 issues message CSQJ105E, an invalid parameter is
  specified which results in abend 0C4-3B in CSQFPMS2.
  

Problem conclusion

• Problem 1:
  New message CSQJ155E is issued when the queue manager opens an
  active LOG dataset that is larger than 4GB.
  
  Problem 2:
  CSQJW207 has been corrected to not pass an invalid parameter
  when issuing message CSQJ105E.
  
  ================================================================
  Add the following new message CSQJ155E to MQ V90x
  IBM knowledge center:
  	- IBM MQ
  	- IBM MQ 9.0.X
  	- IBM MQ
  	- Reference
  	- Diagnostic messages
  	- Messages and reason codes for z/OS
  	- Messages
  
  CSQJ155E
  
      csect-name ACTIVE LOG DSNAME=xxxxx IS LARGER THAN 4GB
  
  Explanation
  
      The queue manager has opened a LOG dataset that has space
  	allocated that is larger than 4GB. The maximum LOG size
  	supported is 4GB.
  
  System action
  
      Processing continues. The LOG dataset continues to be used
  	but the space used is smaller than the space allocated.
  	The allocated highest RBA value overflows a 32bit word and
  	it is truncated.
  	The truncated highest RBA is used resulting in a smaller
  	LOG dataset.
  
  System programmer response
  
      The LOG dataset should be allocated with space not
  	exceeding 4GB.
  	Use Access Method Services to define the data set
  	with a maximum size of 4 GB.
      Refer to "Log dataset definitions"
  	[https://www.ibm.com/support/knowledgecenter/en/SSFKSJ_8.0.0/co
  m.ibm.mq.pla.doc/q006060_.htm]
  	to determine the LOG dataset size.
  
  	To reallocate LOG datasets with space smaller or equal to
  	4GB, carry out these steps:
  
  	1. Allocate and format new COPY1, and COPY2 if used,
  	active log datasets with a maximum size of 4GB, and
         run CSQJUFMT if required.
  	2. Dynamically add the new logs to the queue manager with
  	command DEFINE LOG
  	3. Use command ARCHIVE LOG to roll into the new logs.
  	4. Shutdown the queue-manager when one of the new logs is
  	the current active log dataset.
  	5. Use the CSQJU003 utility to remove the active logs which
  	are too small.  First run CSQJU004 to display the log status
  	in BSDS to make sure the ones to be removed are 'REUSABLE',
  	i.e. they have been successfully archived.
  	6. Restart the queue manager. There should not be any
  	CSQJ115E messages issued.
  

Temporary fix

Comments

APAR Information

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

  PI71899

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

  UI45113 UI45114 UI45115 UI45116 UI45117 UI45118

Modules/Macros

• CSQFJDIC CSQFJDIE CSQFJDIF CSQFJDIK CSQFJDIU CSQFLTXC CSQFLTXE
  CSQFLTXF CSQFLTXK CSQFLTXU CSQFMTXC CSQFMTXE CSQFMTXF CSQFMTXK
  CSQFMTXU CSQJDS02 CSQJW207 N/A
  

Fix information

• Fixed component name

  MQ Z/OS V9

• Fixed component ID

  5655MQ900

Applicable component levels

• R000 PSY UI45113

     UP17/03/30 P F703

• R001 PSY UI45114

     UP17/03/30 P F703

• R002 PSY UI45115

     UP17/03/30 P F703

• R003 PSY UI45116

     UP17/03/30 P F703

• R004 PSY UI45117

     UP17/03/30 P F703

• R005 PSY UI45118

     UP17/03/30 P F703

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.