A storage architecture and method for managing the operation of a network
in a RAID environment is provided in which a storage management agent is
included in each server node of the network. The storage management
agents monitor the status of the drives of the storage array in shared
storage. If a storage management agent identifies a failed drive, the
storage management agent monitors the rebuild of the degraded RAID
volume. During the rebuild of a degraded RAID volume, the storage
management agent determines if a server node has failed, and, if
required, initiates the transfer of the RAID rebuild tasks of the failed
server node to another server node.