摘要 |
PROBLEM TO BE SOLVED: To manage a snapshot without using a dual-redundancy (mirroring) so as to reduce a load on processing and to improve a response. SOLUTION: When the snapshot of a main volume (40) at a desired time is obtained, for example, in an RAID storage system, every time a WR request to any storage region in the main volume (40) is received after the desired time, the data of the storage region are copied to a subvolume (42) (S1, 64) if the data concerned are not yet copied to the subvolume (42) after the desired time, and then the WR processing is executed to the storage region (S2, 60). Further, in case of sequential WR (S3), the data of the storage region to which subsequent sequential WR is forecast to execute are also copied to the subvolume (42) if the data concerned are not yet copied to the subvolume (42) after the desired time. COPYRIGHT: (C)2004,JPO&NCIPI
|