Hitachi, Ltd.

Naoki Higashijima

Ikuya Yagisawa

Yoshihiro Uchiyama

Seiki Morita

A storage control apparatus of the present invention reduces the frequency of disk drive failures. An error management part manages the number of times errors occur in respective disk drives. A disk drive in which the number of errors meets or exceeds a threshold value is selected as a disk drive to be restarted. A restart control part commences difference management prior to restarting the disk drive targeted for restart. A difference management part manages parity group-related update locations using a difference bitmap. After commencing difference management, the restart control part restarts the disk drive in which the error was detected. This makes it possible to resolve an error caused by a firmware hangup or the like.

Storage control apparatus and failure recovery method for storage control apparatus

The storage control device and storage device error control method of the present invention prevent incorrect data from being read out to the host in advance, and thus improve the reliability. For example, the controllerthat controls the disk array subsystem disperses and stores data on respective disksthrough, calculates parity data, and stores this parity data on a disk. In cases where the updating of a certain block (D) fails, so that data is not written into this block, this is an uncorrectable error. The data of this block (D) is recovered on the basis of data and parity data stored on the other disks. The controlleralso recovers data for another block (D) following the block (D) on the basis of other associated data and parity data. As a result, even if by some chance the block (D) which has not suffered an uncorrectable error stores old data, the correct data can be recreated and provided to the host

Storage control device and storage device error control method

A disk array device that prevents the processing of the host device from being interrupted or delayed by drive idle seek, idle seek being executed by all the drives at approximately fixed intervals. The disk array device predicts the approach of the scheduled times to start idle seek of each of the disk drives, . . . and exercises control so that, within the same parity group, idle seek is not executed by two or more drives, . . . at the same time. Within the same parity group, one of the drives whose idle seek scheduled start time is approaching is considered in advance to be executing idle seek and hence data access is denied beforehand. When a data read request with respect to the drivethat is considered to be executing idle seek is received, the required data is recovered by reading data from all the other drives, . . . belonging to the same parity group.

Growing community of inventors

Machida, Japan

Naoki Higashijima