Failover with total loss

If a rack instance is permanently lost, the Remove Member operation removes the rack from the configuration.

Note: To avoid a timeout for the clients that access the file system as an instance is removed from the cluster, in the subsequent steps, before you call the Manage Cluster > Remove Member operation, move the primary, secondary server, cluster manager, and the file system manager to the side that remains active. For more information about the operations to use to make these cluster changes, see Failover active-active operations.

If you have a IBM Storage Scale server instance that was deployed with IBM Storage Scale Pattern 1.2.15.0 or older, run this command from the command line for any of the IBM Storage Scale server nodes to update the cluster and file system manager:

su - gpfsprod -c 'sudo /usr/lpp/mmfs/bin/mmchmgr -c nodeIP' to make the nodeIP the cluster manager.

su - gpfsprod -c 'sudo /usr/lpp/mmfs/bin/mmchmgr fileSystemName nodeIP' to make the nodeIP manager for fileSystemName.

If the rack to be removed hosts the IBM Storage Scale Primary configuration, the IBM Storage Scale Mirror instance becomes the new Primary instance. After the Remove Member operation is performed, you can add another Mirror and Tiebreaker instance to the new Primary instance configuration.

If the Remove Member operation is called on an instance that is not available (the instance cannot be reached by IP address), do not reactivate that instance. If the instance is reactivated, problems can occur that prevent the IBM Storage Scale cluster from functioning properly.

Remember that when creating new IBM Storage Scale Mirror and Tiebreaker configuration instances, you must create disks of the same size so that data can be replicated across the file system.

When creating the initial file system, you can specify shared volumes as you add the Mirror or Tiebreaker instance to the cluster. For subsequent changes to the shared volume configuration, use the Attach Storage Volumes operation for the Mirror and Tiebreaker instances as needed. Specify the file system to use those shared volumes.