Assume that one or more clustered generic services are configured on a Windows Server 2008-based two-node cluster. The Cluster service cannot bring the clustered generic services online if one of the following operations is performed:
A manual attempt to bring the generic service resource online from the Failover Cluster Management administrative console will fail, and the following error message is reported in a message box:Additionally, the following error messages are logged in the cluster log:Notes
- A restart of both nodes in the same timeframe.
- A restart of the cluster service on both nodes in the same timeframe.
- A restart of one node and a restart of the cluster service on the other node in the same timeframe.
- A restart of one node when the other node is offline.
- A restart of the cluster service on one node when the other node is offline.
A manual attempt to bring the generic service resource online from the Failover Cluster Management administrative console will fail, and the following error message is reported in a message box:
The action ‘Bring this resource online’ did not complete. An error occurred while attempting to bring the resource ‘service name’ online. The object already exists.
[RES] Generic Service <Simple Service>: Service 'SimpleService' is controlled by another resource. Error: 5010.
[RHS] RhsCall::Perform_NativeEH: ERROR_OBJECT_ALREADY_EXISTS(5010) because of Resource service name: Open call failed.
- This issue can be triggered by a nonclustered service that is set to start automatically on one or more nodes of the cluster. The nonclustered service interacts with the cluster service by using cluster API.
- This issue also affects clusters that have multiple nodes. The following generic operations will cause this issue to occur when the cluster has multiple nodes (an n-node cluster):
- A restart of all nodes in the same timeframe.
- A restart of the cluster service on all nodes in the same timeframe.
- A restart of i nodes and a restart of the cluster service on the other j nodes, possibly when k nodes are offline, in the same timeframe (where i, j and k are any positive, or zero for k, and n =i+j+k).