Disk array system

A disk array system having a storage device and a storage control device for controlling data storing to the storage device, wherein

• the storage control device has a structure of duplexed clusters having a processor and a memory, and has duplexed power source units that supply power sources to the clusters and a circuit that judges power conditions of the clusters, and
• the storage control device comprises a channel controlling unit equipped with the processor that carries out a communication interface process with other devices, a disk controlling unit equipped with the processor that carries out a communication interface process with the storage device, a cache memory unit for data cache, a shared memory unit for storing condition information, and a connecting unit that connect the units mutually, and
• the processor of the clusters can carry out control to the memory in own cluster, communications with the processor of other cluster, and control to the memory of other cluster, and
• the power source units has a structure having an AC-DC power source unit, and the AC-DC power source unit has a structure including plural switching power sources, and different systems of AC power sources are input to the respective switching power sources to perform AC-DC conversion to output DC power sources to corresponding cluster, and the switching power sources are ones where there is a difference in output stop timings of the DC power sources at power failure owing to differences of power failure durability, and
• in performing control that is executed in response to the case where at least one of the clusters stops operation first at a power failure,
• as the control at power failure, the processor of an operating first cluster blocks a portion of a second cluster that stopped on the basis of blockage judgment, and judges the power source condition of the second cluster on the basis of a signal from the circuit that judges the power condition, and records and stores information showing the power source condition of the second cluster to the shared memory, and
• as the control at power recovery, if the processor of the first cluster recognizes that the portion in blocked condition of the second cluster is in blocked condition owing to the power failure, the processor recovers the portion.