Storage disk failover and replacement system

A storage assembly includes a plurality of multi-unit storage device is provided with a failover procedure that does not require hot-swap capability. Individual units are aggregated into a storage array by an aggregation procedure such as RAID. When a failure occurs data from the failed unit is transferred to a hot spare unit. Also, data from other units is transferred by simulating failure of those units and then the complete device can be removed. In small storage networks all the units may be aggregated as a single array and a spare auxiliary device is summoned and integrated into the array prior to the simulated failure. In larger systems there may be two layers of aggregation and failover is run by the second level of aggregation.