The System Attendant resource cannot fail over in a Microsoft Exchange Server 2003 cluster environment. Additionally, the following events are logged in the Application log on the active node.

Event ID 1069

Event Type: Error
Event Source: ClusSvc
Event Category: Failover Mgr
Event ID: 1069
Date: Date
Time: Time
User: N/A
Computer: Node2
Description:
Cluster resource 'SA' in Resource Group 'ResourceGroupName' failed.

Event ID 1082

Event Type: Erro
Event Source: ClusSvc
Event Category: Checkpoint Mgr
Event ID: 1082
Date: Date
Time: Time
User: N/A
Computer: Node2
Description:
Cluster service failed to restore a registry key for resource SA when it was brought online. This error code was 6001.

Event Type: Error
Event Source: ClusSvc
Event Category: Checkpoint Mgr
Event ID: 1082
Date: Date
Time: Time
User: N/A
Computer: Node2
Description:
Cluster service failed to restore a registry key for resource SA when it was brought online. This error code was 6001.

Event Type: Error
Event Source: ClusSvc
Event Category: Checkpoint Mgr
Event ID: 1082
Date: Date
Time: Time
User: N/A
Computer: Node2
Description:
Cluster service failed to restore a registry key for resource SA when it was brought online. This error code was 6001.

Event Type: Error
Event Source: ClusSvc
Event Category: Checkpoint Mgr
Event ID: 1082
Date: Date
Time: Time
User: N/A
Computer: Node2
Description:
Cluster service failed to restore a registry key for resource SA when it was brought online. This error code was 1917. Some changes may be lost.

Event ID 1121

Event Type: Error
Event Source: ClusSvc
Event Category: Checkpoint Mgr
Event ID: 1121
Date: Date
Time: Time
User: N/A
Computer: Node1
Description:
The crypto checkpoint for cluster resource 'SA' could not be restored to the container name 'ContainerName'. The resource may not function correctly.

This problem occurs if two nodes use different accounts to run the Cluster Service.

After you verify that different accounts are configured for the Cluster Service, configure the nodes to use the same account for the Cluster Service. To do this, follow these steps.

Note The following steps assume that you are running a typical two-node cluster.

1. Stop the Cluster Service on node 1.
2. Change the startup type of the Cluster Service to Manual.
3. Wait until all resources are listed under node 2 in Cluster Administrator.
4. Change the account of the Cluster Service on node 1.
5. Change the startup type of the Cluster Service back to Automatic.
6. Start the Cluster Service on node 1.
7. Test a failover between node 1 and node 2.