| Legal Information |
|
If you simultaneously lose all of the nodes of an Exchange 2000 cluster, you must recover the entire cluster. The process for recovering an entire cluster includes many of the same procedures for recovering stand-alone Exchange 2000 member servers. For more information about stand-alone server recovery methods, see "Selecting an Exchange 2000 Disaster Recovery Strategy."
If you do not have the required full computer backups or Windows backups of the nodes in your cluster, you may still be able to recover your entire cluster. To attempt this type of recovery, you must have backups of your Exchange databases (or your Exchange database files and transaction log files must be intact on one of your cluster's shared disk resources). You must also have sufficient informational records about your cluster configuration. For information about how to record cluster information, see "Maintaining Informational Records About Your Clusters".
Important To rebuild an entire cluster using your cluster's information records and Exchange database backups, contact Microsoft Product Support Services at http://support.microsoft.com. The procedures required in this type of recovery are for advanced-level administrators only. Furthermore, advanced-level administrators should only consider this cluster recovery method if there is no alternate method available.
When implementing a recovery strategy for an entire cluster, the first node you recover (also known as "first node") must be the node that owned the quorum disk resource at the time you created the backup sets for your nodes. After you recover the first node, ensure that all of your cluster resources come online. After your cluster resources are online, you can insert new nodes, insert stand-by recovery nodes, or continue to restore or rebuild additional failed nodes.
The following information provides further explanation about using the "restore the server" and the "repair the server" methods to recover the first node of your cluster:
If you use the "restore the server" method to recover the first node, you may need to restore one or more of the cluster's shared disk resources (for example, the quorum disk resource or Exchange databases) after you restore the full computer backup set. For more information about how to restore these shared disk resources, see "Restoring Shared Disk Resources". After you ensure that the first node in the cluster is able to bring all cluster resources online, you can insert new nodes, insert stand-by recovery nodes, or continue to restore nodes from your full computer backup sets.
If you use the "rebuild the server" method to recover the first node, you must recover the quorum disk resource (if necessary) after you restore the Windows backup set. After you restore your Windows backup set, the Cluster service starts, and the objects in your cluster should appear as they did prior to the disaster. To view your cluster information, use Cluster Administrator.
After you install Exchange 2000 in Install mode, restore your Exchange database backups (if applicable). After you ensure that the first node in your cluster is able to bring all cluster resources online, you can insert new nodes, insert stand-by recovery nodes, or continue to rebuild nodes from your backup sets.
Important When you install Exchange to a cluster node as part of a cluster node recovery, you must run Exchange 2000 Setup in Install mode, not in Disaster Recovery mode.
For general information about these recovery methods, including the advantages and disadvantages of each, see "Selecting an Exchange 2000 Disaster Recovery Strategy."
To view tables that contain various disaster recovery backup requirements and restore procedures, see "Disaster Recovery Tables."
The information in this article applies to:
| Search Knowledge Base | Feedback |