摘要 |
PROBLEM TO BE SOLVED: To suppress a factor in causing the loss of data in a distributed storage system. SOLUTION: Upon receiving check instructions of data 3a in a storage device 3, a data checking means 1c reads the data 3a from the storage device 3 and determines whether the data 3a is normal or not. When abnormality of data is detected by the data checking means 1c, management information is referred to by a data acquisition means 1d to determine a duplexed counterpart storage node storing redundant data corresponding to the abnormality detected data, and the redundant data corresponding to the abnormality detected data is acquired from the duplexed counterpart storage node. The redundant data acquired by the data acquisition means 1d is written into a storage area of the data 3a in the storage device 3 by overwriting. COPYRIGHT: (C)2008,JPO&INPIT
|