Restoring the cluster quorum
The cluster quorum is backed up as part of System State or Shadow Copy Components for each node. To restore the cluster quorum, other nodes in the cluster must be offline. If you cannot take the other nodes in the cluster offline before you restore the cluster quorum, follow the instructions in the next procedure.
To force the restore of the cluster quorum when other nodes are online
- Create the following registry entry:
HKEY_LOCAL_MACHINE\Software\VERITAS\NetBackup\CurrentVersion\ Config\Cluster_Database_Force_Restore
The actual registry entry must be on a single line. If required, create this registry value as data type REG_SZ with a value of YES. When this value is set, the cluster service is stopped for any nodes that are online.
- If Active Directory is running on the target server or the target server is currently a domain controller, the restore procedure is different.
To restore the cluster quorum
- If the other nodes in the cluster are online, create the registry entry as described in the previous procedure.
- In the NetBackup Backup, Archive, and Restore interface, open a Restore window.
- Select the backup image (or range of images) that contains the backup of the System State or Shadow Copy Components for this node of the cluster.
- In the All Folders pane, select System_State or Shadow Copy Components.
- If the disk where the cluster quorum previously resided is changed, set the registry entry. This action forces the restore to continue.
The disk may have been replaced with a new one. Or the disk configuration may have been changed so that the cluster quorum now resides on a different disk. In the latter case, the drive letter can remain the same for the disk that contained the cluster quorum. The drive letter can remain the same even if the configuration has changed. It can also remain the same if the disk signatures in the restore media do not match the signatures in the cluster quorum.
- Continue the restore.
- When the restore operation is complete, restart the cluster node. Or use the Failover Cluster Manager to restart the cluster service on any nodes where it was stopped.