PI12800: SOS DUE TO IE_BUFF SUBPOOL USAGE IN ESDSA FOR ECI REQUEST WHEN CICS SOCKET RECEIVE TIMES OUT

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• Your CICS region shows a high usage of ESDSA.  Looking further
  you notice this is for subpool IE_BUFF.  There are a number of
  DFHIE1104 Error attempting socket receive from ECI client.
  The dump shows only one in progress ECI based task.  The trace
  shows several ECI tasks starting and completing successfully
  and managing their IE_BUFFs correctly.
  
  There seems to have been a network error leading CICS to only
  receive part of an ECI request and the CICS socket receive
  gets timed out. Once that timeout happens the CIEP task ends up
  in its server_error routine.  CICS will finish the CIEP task
  without freeing the IE_BUFF and also without issuing another
  socket receive.  This means that the IE_BUFF and IECCB will
  not get freed. The orphaned IE_BUFF and IECCB storage will
  remain until CICS is restarted.  If there are enough network
  errors then CICS could eventually end up short on storage.
  
  
  Additional Symptom(s) Search Keyword(s): KIXREVSWM
  DFHSM0133  SOS Short on storage
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: All                                          *
  ****************************************************************
  * PROBLEM DESCRIPTION: Short on storage in ESDSA due to high   *
  *                      storage usage in subpool IE_BUFF        *
  *                      together with a large number of         *
  *                      DFHIE1104 messages.                     *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  A network error leads CICS to only receive part of an ECI
  request and then the CICS socket receive gets timed out
  and a DFHIE1104 message is issued:
  
  DFHIE1104 Error attempting socket receive from ECI client.
  
  Once that timeout happens the CIEP task ends up in its
  SERVER_ERROR routine and CICS finishes the CIEP task without
  freeing the IE_BUFF and also without issuing another socket
  receive.  The IE_BUFF and IECSB storage associated with the
  failed task will remain until CICS is restarted.
  
  If there are enough of these network errors CICS will eventually
  go short on storage in ESDSA due to high storage usage in
  subpool IE_BUFF.
  

Problem conclusion

• The SERVER_ERROR routine called by a CIEP task has been changed
  to free IE_BUFF when an ECI synchronous receive request is timed
  out.
  

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:

  PI20180 UI19525

Modules/Macros

• DFHEISO  DFHIEDM  DFHIEDUF DFHIEIE  DFHIETRI DFHIEXM
  

Fix information

• Fixed component name

  CICS TS Z/OS V4

• Fixed component ID

  5655S9700

Applicable component levels

• R700 PSY UI19525

     UP14/07/17 P F407

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.