Error Message:

Cluster disk resource 'name ' did not respond to a SCSI maintenance command.

Explanation:

A SCSI maintenance command is a reserve operation (a write operation) to the quorum disk, which 1) maintains ownership of the disk and 2) makes sure the disk is still there. If the reserve operation fails, either the disk is broken or (more likely) the cluster's other node has taken ownership of the disk. Failure of the node that owns the quorum device to send SCSI maintenance commands can happen if the node is in kernel debugging mode, or because of software or hardware problems. Assume that node A owns the quorum device. In normal cluster operations, the node that does not own the quorum device (node B) periodically challenges node A's ownership of the quorum, then after an interval follows up with a check. Between node B's challenge and check, node A must send its SCSI maintenance command in order to retain ownership of the quorum device. If node A is sluggish or comatose and fails to send the maintenance command, node B's check succeeds, and B takes ownership of the quorum. Node A shuts down, and this error message results.

User Action:

The node that was sluggish or comatose must rejoin the cluster. If the cause of the node's sluggishness is software, call Microsoft Product Support Services for further information. If the sluggishness is due to the node being in kernel debugger mode, take it out of the debugger. If the cause is hardware, troubleshoot the hardware.