IJ15739: clstrmgr generates core after several dare operations

PowerHA-7.2.3 SP1

APAR status

• Closed as fixed if next.

Error description

• This apar fixes some clstrmgr memory leaks.
  The larger the cluster configuration (number of
  RGs and resources) the more likely the cluster
  will be affected. A possible stack seen in errpt
  is as follows:
   ADDITIONAL INFORMATION
    pthread_k AC
    ??
    _p_raise 4C
    raise 48
    abort BC
    die__Fi 670
    getODM__8 15C
    initializ 6C
    RGPA_Perf 5C
    clrgpa__1 494
    clrgpa__F 64
    clipc_set 258
    IpcHandle 590
    DoMainLoo 73C
    main 944
    __start 6C
  

Local fix

Problem summary

• *
    * USERS AFFECTED:
    * Systems running PowerHA 7.2.3 with
    * cluster.es.server.rte at the 7.2.3.0 level.
  *
    * PROBLEM DESCRIPTION:
  * In a PowerHA SystemMirror cluster, an unexpected halt of the
    * primary node and RG failover may occur.
    * This happens when the clstrmgr daemon leaks memory and core
    * dumps when heap memory is exhausted.
    * The problem can occur after a PowerHA cluster running the
    * affected levels below has been running for a long time, and
    * clstrmgr daemon's memory usage has grown very large.
  *
    * RECOMMENDATION:
    * Install APAR IV15739.
    * Prior to fix availability, an interim fix is available from
    * either
    * ftp://aix.software.ibm.com/aix/ifixes/ij15739/
    * https://aix.software.ibm.com/aix/ifixes/ij15739/
  *
  

Problem conclusion

Temporary fix

•   *********
    * HIPER *
    *********
  

Comments

APAR Information

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

  IJ09170

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

Fix information

• Fixed component name

  POWERHA SYSMIR

• Fixed component ID

  5765H3900

Applicable component levels