OA34152: TEMS BECOMES UNRESPONSIVE DUE TO LOCK DELAY.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• APAR Type:Field
  Approver Initials: LK
  
  Severity:3
  Reported Release: 610
  Compid: 5724K0900 TEMS on Distributed Platform
  
  ABSTRACT:TEMS becomes unresponsive due to lock delay.
  
  PROBLEM DESCRIPTION:
  The customer encounters the problem that all agents immediately
  go OFF-LINE, and NEVER be back ON-LINE unless the customer
  reboots of TEMS server.   After rebooting TEMS AIX box and
  starting TEMS, all agents come back ON-LINE.
  
  The problem occasionally happens on the customer's system about
  once a month or two months.
  
  TEMS becomes unresponsive due to lock delay.
  
  ENVIRONMENT:
  ITM 6.1 FP07 IF0004 / AIX 5.3
  
  DOCUMENTS:
  Trace Log <full filename/Location>:
  File Name: 68946.6X4.760.2009118a.tar.Z
  Directory: /ecurep/pmr/6/8/68946,6X4,760
  
  Files are contained:
  - Agent_logs/* : agents' logs of Linux, NT, AIX platform
  - TEMS_logs/logs : TEMS RAS1 traces with KDC_DEBUG=Y and
  KDE_DEBUG=Y
  - TEMS_logs/procstack_kdsmain.1.out : Output of procstack
  - TEMS_logs/snapcore_745576.pax.Z : Output of snapcore
  - listsystems_after_reboot.out : all agent list
  - off-line_agents_Nov15.out : MS_Offline agent on Nov 15.
  

Local fix

• No workaround available except to recycle TEMS.
  

Problem summary

• ****************************************************************
  * USERS AFFECTED: All TEMS users.                              *
  ****************************************************************
  * PROBLEM DESCRIPTION: OCCASIONALLY, AGENTS THAT ARE KNOWN TO  *
  *                      BE ON-LINE ARE REPORTED AS BEING OFF-   *
  *                      LINE.                                   *
  ****************************************************************
  * RECOMMENDATION: Apply this ptf.                              *
  ****************************************************************
  The problem is a result of the management server
  experiencing a thread deadlock condition in the management
  agent proxy.  This usually occurs after long periods of
  management server uptime without a restart.
  

Problem conclusion

• The code was modified so that the lock pool size is now
  configurable using the environment variable
  MAX_CTIRA_RECURSIVE_LOCKS.  The default value is still 20
  and the maximum can be increased up to 200.  The variable
  can be set in the management server environment file.
  Typically, this value should be increased under the
  direction of IBM customer support.
  

Temporary fix

Comments

APAR Information

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

  IZ76288

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

Modules/Macros

• KDSINDFE
  

Fix information

• Fixed component name

  MGMT SERVER DS

• Fixed component ID

  5608A2800

Applicable component levels

• R622 PSY UA56912

     UP10/10/02 P F010

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.