IV65639: CAA PARTITIONED CLUSTER CAUSED BY WAIT_ON_NODE_BRINGUP ERROR APPLIES TO AIX 6100-09

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• ****************************************************************
  * USERS AFFECTED:
  * Systems running the 6100-09 Technology Level with
  * bos.cluster.rte below the 6.1.9.45 level.
  ****************************************************************
  * PROBLEM DESCRIPTION:
  * A possible split-brain issue was introduced for
  * PowerHA 7.1 customers.  The issue causes the disk
  * heartbeat signals to be delayed by ~10 seconds during
  * cluster startup.  If network heartbeat is also missed
  * during the time window, the HA node that was just
  * rebooted would miss any other Cluster nodes that are
  * already up. This would result in the rebooted node
  * declaring an independent one node Cluster causing the
  * split.  This apar ensures that the repository disk
  * heartbeat is enabled earlier during the node startup and
  * also includes further safeguards for the node cluster
  * join sequence.
  ****************************************************************
  * RECOMMENDATION:
  * Install APAR IV65639.
  * Prior to fix availability, an interim fix is available from
  * either
  * ftp://aix.software.ibm.com/aix/ifixes/iv65639/
  * https://aix.software.ibm.com/aix/ifixes/iv65639/
  ****************************************************************
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED:
  * Systems running the 6100-09 Technology Level with
  * bos.cluster.rte below the 6.1.9.45 level.
  ****************************************************************
  * PROBLEM DESCRIPTION:
  * A possible split-brain issue was introduced for
  * PowerHA 7.1 customers.  The issue causes the disk
  * heartbeat signals to be delayed by  10 seconds during
  * cluster startup.  If network heartbeat is also missed
  * during the time window, the HA node that was just
  * rebooted would miss any other Cluster nodes that are
  * already up. This would result in the rebooted node
  * declaring an independent one node Cluster causing the
  * split.  This apar ensures that the repository disk
  * heartbeat is enabled earlier during the node startup and
  * also includes further safeguards for the node cluster
  * join sequence.
  ****************************************************************
  * RECOMMENDATION:
  * Install APAR IV65639.
  * Prior to fix availability, an interim fix is available from
  * either
  * ftp://aix.software.ibm.com/aix/ifixes/iv65639/
  * https://aix.software.ibm.com/aix/ifixes/iv65639/
  ****************************************************************
  

Problem conclusion

• Ensure that the repository disk heartbeat is enabled earlier
  during the node startup and also include further safeguards for
   the node cluster join sequence.
  

Temporary fix

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

Comments

• 6100-08 - use AIX APAR IV65638
  6100-09 - use AIX APAR IV65639
  6100-09 - use AIX APAR IV65639
  6100-09 - use AIX APAR IV65639
  7100-02 - use AIX APAR IV65643
  7100-03 - use AIX APAR IV65472
  7100-03 - use AIX APAR IV65472
  

APAR Information

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

  IV65472

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

Fix information

• Fixed component name

  AIX 610 STD EDI

• Fixed component ID

  5765G6200

Applicable component levels

• R610 PSY U861492

     UP15/11/20 I 1000

PTF to Fileset Mapping

• U865869 bos.cluster.rte 6.1.9.45

• U861492 bos.cluster.rte 6.1.9.100