Sometimes, cluster member with the Highest Priority is not able to become Active member after installing the policy in the following scenarios:

• In cluster with 2 members:

  1. "Maintain current active Cluster Member" is enabled.
  2. Freeze Mechanism is enabled (refer to sk32488 - value of kernel parameter fwha_freeze_state_machine_timeout is other than 0).
  3. Members' Priorities are changed (e.g., priority of Member_B is increased; priority of Member_A is decreased).
  4. After policy installation, the original member that had the Highest Priority (Member_A) is still Active.

• In cluster with 3 members:

  1. "Maintain current active Cluster Member" is enabled.
  2. Freeze Mechanism is disabled (refer to sk32488 - value of kernel parameter fwha_freeze_state_machine_timeout is set to 0).
  3. Members' Priorities are changed (e.g., priority of Member_B is increased; priority of Member_A is decreased).
  4. After policy installation, the original member that had the Highest Priority (Member_A) is still Active.

/var/log/messages file on the cluster members shows that during policy installation, the cluster members whose priority was changed exit the Freeze State, although those members were not in Freeze State.

In cluster with 3 members with enabled Freeze Mechanism, the behavior is correct:

1. "Maintain current active Cluster Member" is enabled.
2. Freeze Mechanism is enabled (refer to sk32488 - value of kernel parameter fwha_freeze_state_machine_timeout is other than 0).
3. Members' Priorities are changed (e.g., priority of Member_B is increased; priority of Member_A is decreased).
4. After policy installation, the member with the Highest Priority (Member_B) becomes the new Active.