PH50082: CSQY228E ACELIM REACHED: S0DC SLIP DUMP: 5C6-00E2000C ABEND IN CSQSFBK

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A 5C6-00E2000C abend occurred during an attempt to get ECSA
  storage. As ACELIM was configured, CSQY228E ACELIM REACHED also
  occurred. The MQ MSTR job was one of the high users of ECSA due
  high usage of MQ thread-related subpools such as RAL,
  "ccb for ssam", "conn,agnt,fvl+others", ROB, and ACE/PEB. There
  was a buildup of MMCLTRX system threads. These tasks were all
  waiting to obtain a latch.
  
  The latch holder was waiting for a latch that was held by a
  thread for an application job. The application job had received
  an earlier ABEND07D RC0 (Functional Recovery Routine (FRR)
  stack full), which was followed by an ABEND S0DC - REASON
  CODE 08 to request a MEMTERM of the address space. A thread in
  the address space still held a latch. z/OS APAR OA62248 assists
  in obtaining documentation for the ABEND0DC.
  
  MQ development has been able to recreate the problem whereby a
  task with an FRR stack depth of 7 calls MQ for RRS commit
  processing and subsequently abends S07D. This APAR is to
  investigate how best to address this problem.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All users of IBM MQ for z/OS Version 9       *
  *                 Release 2 Modification 0 and                 *
  *                 Release 3 Modification 0.                    *
  ****************************************************************
  * PROBLEM DESCRIPTION: An 07D-00 abend is seen in the Queue    *
  *                      Manager address space leading to an     *
  *                      0DC-08 memterm being scheduled for the  *
  *                      connected application address space.    *
  ****************************************************************
  An 07D-00 abend is seen in the Queue Manager address space due
  to a failed attempt to add an additional FRR to a full FRR
  stack. This leads to a 0DC-08 memterm being scheduled for the
  connected application address space.
  
  The FRR stack was unusually full due to a series of rare code
  paths all being traversed on the same task. In this case the FRR
  stack held entries from DB2 stored procedures, RRS commit
  processing, MQ Logging processing, a checkpoint being initiated
  due to LOGLOAD being hit and latch contention.
  

Problem conclusion

• The code has been altered so that additional FRR entries are not
  required as part of initiating a checkpoint when Logload has
  been hit. This has the affect of reducing the current depth of
  the FRR stack avoiding the reported symptoms of this APAR.
  
  Additionally a back stop abend 5C6-00E513A1 has been introduced
  to prevent attempts to add an FFR to the FRR stack when there is
  not enough space to do so. This will help identify other
  potential paths to reach the symptoms described by this APAR.
  Instead of an application address space being memtermed this new
  abend will be issued and a dump produced that can be given to
  IBM support for analysis.
  

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:

  UI94305 UI94306

Modules/Macros

• CSQRCRQS CSQRCRSC CSQRCSHT CSQRIT01 CSQVSRX
  

Fix information

• Fixed component name

  IBM MQ Z/OS V9

• Fixed component ID

  5655MQ900

Applicable component levels

• R200 PSY UI94306

     UP23/11/15 P F311 ¢

• R300 PSY UI94305

     UP23/11/15 P F311 ¢

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.