Start of change

6027-4210 [E] Paxos state on tiebreaker disk diskName contains wrong cluster ID. Expected: clusterId is: clusterId

Explanation

When the cluster is configured with tiebreaker disks, the CCR stores its Paxos state on a reserved area on these tiebreaker disks. The tiebreaker disks are also used by IBM Storage Scale to declare quorum , if enough quorum nodes are not reachable. This message is returned by the CCR, if it found a Paxos state on a tiebreaker disk with an unexpected cluster identification number. This means that the Paxos state comes from another cluster. Possible reason for this failure: The affected tiebreaker disks are replicated or mirrored from another cluster, typically in an active-passive disaster recovery setup.

User response

By reconfiguring the affected tiebreaker disks, the wrong Paxos state will be overwritten with a correct state. Perform the following procedures:
  1. Ensure that all quorum nodes are up and IBM Storage Scale reached quorum.
  2. Reconfigure the cluster not to use the affected disks as tiebreaker disks, by using the "mmchconfig tiebreakerDisks" command.
  3. Reconfigure the cluster again, now to use the affected disks as tiebreaker disks by using the "mmchconfig tiebreakerDisks" command.
End of change