Restoring Exchange Clusters


The disaster recovery processes for restoring Exchange clusters are similar to the processes for restoring data on stand-alone Exchange servers. However, before you start to perform recovery processes on your clusters, it is helpful to understand how Exchange cluster resources can continue to remain online, even if one of the nodes experiences a failure.

If one of the nodes in a cluster fails (known as a failover event), the Cluster service takes control of the cluster. Following the failover, one of the possible owner nodes for the resource group tries to take control of that group. If all the resources can come online for the new node, that node continues to perform the tasks that were previously performed by the damaged node. If the resources cannot come online for the new node, that node will fail over to the next node. This process continues until all possible owner nodes for that group cannot come online.

Similarly, if one of the resources in an Exchange Virtual Server (EVS) fails, the EVS goes offline. A possible owner node tries to start all the resources for that EVS. If the resources cannot come online for the new node, that new node fails over to the next possible owner node. If all nodes cannot bring the EVS resources online, the resources on that virtual server will be unavailable to Exchange clients until the problem is resolved.

An important difference in disaster recovery processes for Exchange clusters is the task of identifying what caused a particular resource to fail. If a problem occurs, first determine whether the failure is on a single node (which indicates that there are problems with the node's files) or on every node (which indicates that there are problems with the cluster's objects or the shared cluster resources). To determine the cause of the failure, search the event logs in Event Viewer. You can also search for solutions in the Microsoft Knowledge Base.

If you still cannot determine the cause of the failure, you can perform the repair options listed in "Repairing Windows Server 2003" or "Repairing Exchange Server 2003." If you cannot repair the node or the whole cluster, consider replacing the node or recovering the node, cluster, or resources (such as the quorum disk resource or mailbox and public folder stores).

This section provides the following procedural information about restoring Exchange clusters:

  • Replacing damaged Exchange cluster nodes.

  • Restoring or rebuilding a cluster node from backups.

  • Restoring shared disk resources.

  • Recovering a whole Exchange cluster.