Repairing a Lost Sun StorEdge A5000 Connection

When a connection from a disk enclosure to one of the cluster nodes fails, the failure is probably due to a bad SCSI-2 cable or an SBus card.

In any event, the node on which the failure occurred will begin generating errors when the failure is discovered. Later accesses to the disk enclosure will generate additional errors. The node will exhibit the same behavior as though power had been lost to the disk enclosure. I/O operations from the other nodes in the cluster are unaffected by this type of failure.

To diagnose the failure, use the procedures for testing the card module in the service manual for your Sun Cluster node to determine which component failed. You should free up one node and the disk enclosure that appears to be down, for hardware debugging.

How to Repair a Lost Sun StorEdge A5000 Connection

1. Prepare the Sun Cluster system for component replacement.

   Depending on the cause of the connection loss, prepare the Sun Cluster node with one of the following procedures.

   • If the failed component is an SBus FC-100 host adapter, see Chapter 7, Administering Server Components, to prepare the Sun Cluster node for power down.

   • If the problem is a bad FC-100 fiber optic cable, the volume management software will have detected the problem and prepared the system for cable replacement.

2. Replace the failed component.

   If the FC-100 fiber optic cable or SBus FC-100 host adapter fails, refer to the Sun StorEdge A5000 Installation and Service Manual for detailed instructions on replacing them.

3. Recover from volume management software errors.

   Use the procedures described in "Recovering From Power Loss".

This completes the procedure for repairing a lost connection.