In the event of a failure of the write server’s machine, it is possible to recover to the point of the failure. This requires obtaining access to the write server’s disk resources from a working machine.
If the write server can be repaired and brought back up with all the database resources operational, then it is only necessary to restart the write server and the dbremote processes.
If the write server cannot be repaired, you may still be able to reconstitute a working write server by reconnecting the failed machine’s disk resources to a new machine. You can accomplish this by recabling disk drives, or by reconfiguring the logical layout of a shared disk-storage subsystem.
