Dell 4200 manual Troubleshooting

Table 5-1. Troubleshooting (continued)

Problem	Probable Cause	Corrective Action
One or more of the SCSI control-	The controllers have con-	Change one of the controller SCSI IDs so that the
lers are not detected by the	flicting SCSI IDs.	ID numbers do not conflict. The controller in the
system.		primary node should be set to SCSI ID 7, and the
		controller in the secondary node should be set to
		SCSI ID 10. Refer to Chapter 3 for instructions
		for setting the SCSI IDs on the nodes.

One of the nodes can access one of the shared hard-disk drives, but the second node cannot.

The drive letters assigned to the hard-disk drive differ between the nodes.

The SDS 100 storage system has not been upgraded with the cluster-specific firm- ware.

Change the drive letter designation for the shared hard-disk drive so that it is identical in all nodes.

Ensure that the SMB-connected node on the clus- ter is running the cluster-specific firmware. Upgrade the SDS 100 firmware by powering down the cluster and then starting it up again. During start-up, the cluster-specific firmware on the node checks the version of the SDS 100 firm- ware. If the SDS 100 is found to be running the wrong version of firmware, the node proceeds to upgrade it automatically with the correct firm- ware version.

The SCSI cable between the node and the shared storage subsystem is faulty or not connected.

Attach or replace the SCSI cable between the cluster node and the shared storage subsystem.

Server management functions are unavailable when both nodes are functional.

The SMB cable is not con- nected properly to the SDS 100 storage system(s).

Check the SMB connections. The primary node should be connected to the first storage system, and the second storage system (if present) should be connected to the first storage system. Refer to Chapter 2 for information about connecting the SMB cable.

Clients are dropping off of the network while the cluster is fail- ing over.

The service provided by the recovery group becomes temporarily unavailable to clients during fail-over. Clients may lose their con- nection if their attempts to reconnect to the cluster are too infrequent or if they end too soon.

Reconfigure the dropped client to make longer and more frequent attempts to reconnect back to the cluster.

Troubleshooting 5-3