PM45737: AUXILIARY TEMPORARY STORAGE OUT OF SPACE FOLLOWED BY MULTIPLE EXCEPTION MONITORING RECORDS

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• A region has suffered an out of space condition on the auxiliary
  temporary storage dataset. Right after this condition occurred
  there are multiple exception monitoring records being cut.
  
  Upon further review of CICS Internal Trace shows a task being
  purged and all the tasks waiting for AUX_SPACE are given the
  chance to run again. This is the design mechanism that
  temporary storage uses when giving other tasks the chance to
  run. The logic in DFHTSWQ will make an exception data put call
  to monitoring domain, and this is in turn visible in the trace
  entries for APEX for XMNOUT.
  
  Giving every task the chance to run again allows the CICS
  dispatcher to honor dispatching priorities when allocating which
  task gets the chance to run next and get the resource in
  question. Normally there are relatively few waiters and the
  effect of resuming them all does not have that great of impact
  on a system.
  
  For this instance the logic path for recovery following a
  timeout on a write invokes the same method as used for basic
  delete processing. In the case of a queue deletion it makes
  sense to resume all waiters for storage on DFHTEMP. But in the
  case of a write that has timed out it will not have freed up
  storage on DFHTEMP. In this particular case, there is no benefit
  to resuming any waiters, and doing so would indeed lead to
  unfavorable conditions.
  
  Additional Symptom(s) Search Keyword(s):
  KIXREVRHC
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All                                          *
  ****************************************************************
  * PROBLEM DESCRIPTION: Unnecessary resumes of tasks waiting    *
  *                      for TSAUX storage.                      *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  When a writeq times out or is purged while waiting for auxiliary
  temporary storage to become available, routine tsaux_delete will
  resume all waiters for TSAUX. These tasks are redispatched, but
  will typically fail to find any freed-up space on DFHTEMP, so
  have to resuspend on TSAUX once more. This leads to churn within
  CICS and unnecessary CPU being used.
  

Problem conclusion

• DFHTSQR has been changed to only resume waiters if space has
  been freed up as part of the tsaux_delete processing.
  

Temporary fix

• FIX AVAILABLE BY PTF ONLY
  

Comments

APAR Information

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

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

  PM48753 UK73856

Modules/Macros

•    DESTSAM  DESTSQR  DFHTSAM  DFHTSAMC DFHTSPT
  DFHTSQR
  

Fix information

• Fixed component name

  CICSTS V3 Z/OS

• Fixed component ID

  5655M1500

Applicable component levels

• R500 PSY UK73856

     UP11/11/19 P F111

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.