PH02959: HIGH PRIVATE STORAGE USAGE DUE TO K&T NOTIFICATIONS

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Queue managers in a QSG can exhibit high storage usage following
  a large number of K&T notifications from other queue managers in
  the QSG.
  .
  .
  Information from PH04105 that was closed as a duplicate of this
  APAR:
  Abstract: OPTIMIZE USE OF MQ'S DXWB CONTROL BLOCKS TO PROVIDE
            STORAGE CONSTRAINT RELIEF IN SUBPOOL 229 KEY 7
  .
  DXWB control blocks for MQ represent an item of work from XCF
  that needs to be processed by MQ.
  
  There can be situations, such as a problem in XCF or DB2 or a
  problem with a looping or high volume application that can
  cause higher use of these control blocks.
  
  L3 has found a way to optimize use of the DXWBs to avoid or
  delay running out of storage related to these blocks.
  
  
  Additional Symptom(s) Search Keyword(s):
  
  Common symptoms associated with a high number of DXWB and EKTM
  control blocks:
  
  - The amount of free storage reported in CSQY220I decreases,
    perhaps suddenly, and the message becomes more frequent:
      CSQY220I Queue manager is using mm MB of local storage, nn
               MB are free
  
  - Eventually storage may run out:
      CSQY221I Queue manager is short of local storage
      CSQY222E Queue manager is critically short of local storage
               - take action
  
  - ABENDS5C6 occurs with reason codes such as 00E2000B
  
  - ABEND878-00000010
  
  Queue Sharing Group QSG     Kiss and Tell
  878 S878 S0878 ABEND878 ABENDS878
  Subpool 229 Key 7   SP229 KEY7 K7  private storage
  

Local fix

• - For BATCH and RRSBATCH applications, consider setting
    CONNSWAP=NO in CSQ6SYSP to prevent MQ applications from
    being swapped out in the middle of an MQI call. This will
    avoid contention that may lead to a build-up of the
    control blocks.
  
  - Search for APARs relating to DXWB blocks:
    https://ibm.biz/BdXXKn
  
  - Check the syslog for errors with XCF (IXLnnnn), DB2
    (DSNnnnn), IMS (DFSnnnn), or MQ (CSQnnnn) that may prevent
    normal operation. For instance, the thread in MQ that
    processes the DXWBs may be hung due to a problem in DB2:
      DSNJ111E OUT OF SPACE IN ACTIVE LOG DATA SETS
  
  - Check logrec to see if any prior abends may have interrupted
    normal processing. See sample JCL at
   http://www.ibm.com/support/docview.wss?rs=171&uid=swg21291812
  
  - Review whether a looping application may be causing the
    Coupling Facility (CF) to be flooded with requests
  .
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of IBM MQ for z/OS Version 8       *
  *                 Release 0 Modification 0.                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: Abend S878, storage exhaustion and      *
  *                      abnormal queue manager termination can  *
  *                      occur when a large number of XCF        *
  *                      messages are received from other        *
  *                      queue managers in the queue sharing     *
  *                      group.                                  *
  ****************************************************************
  When messages arrive over XCF, the queue managers XCF
  message/group exits obtain a DXWB control block to contain the
  message, and queue it for processing by the XCF worker task.
  If a large number of messages arrive (for example due to
  messages arriving on shared queues with TRIGTYPE(EVERY) ),
  the number of DXWBs waiting to be processed grows, and the
  storage allocated for these control blocks can result in private
  storage shortages and S878 abends.
  This is exacerbated by unoptimal use of this storage, leading to
  more storage being allocated than is required, and storage
  fragmentation.
  

Problem conclusion

• The use of DXWBs is changed to use more appropriately sized
  storage areas, reducing the storage requirement for this
  processing.
  

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:

  PH03837 UI59655

Modules/Macros

• CSQ5GRX0 CSQ5MGX0 CSQ5RCMQ CSQ5STRT CSQ5XWK1 CSQSIQCP
  

Fix information

• Fixed component name

  IBM MQ Z/OS V8

• Fixed component ID

  5655W9700

Applicable component levels

• R000 PSY UI59655

     UP18/12/14 P F812 ¢

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.