Method and system for distributing and reestablishing data of magnetic disc array

Abstract

The invention relates to a method and a system for distributing and reestablishing data of a magnetic disc array. The magnetic disc array comprises a magnetic disc array controller, a normal disc and a hot standby disc, wherein the magnetic disc array controller is in charge of distribution of the data on the normal disc and the reestablishment of the magnetic disc array after the fault of one or more magnetic discs, and the magnetic disc array controller selects the hot standby disc for reestablishment after the fault of one or more magnetic discs; the storage space of the hot standby disc is divided into alternatively distributed data spaces and idle spaces; the data with size of an I / O (input / output) heat statistic storage area are read and stored in a main magnetic disc, an operation result is written into the data space of the hot standby disc after the data are subjected to XOR operation, in the operation of finishing primary reestablishment, track seeking distance of an I / O magnetic head on the hot standby disc is reduced and the reestablishment speed of the magnetic disc array is accelerated.

Description

technical field [0001] The technical field of the present invention is the field of information storage, in particular to a disk array data distribution and reconstruction method and system. Background technique [0002] A disk array is a storage system that organizes multiple physical disks through a certain data distribution method to provide data read and write services through concurrent I / O. Disks are mechanical components that have a very high probability of failure. When a disk in the disk array fails, the disk array enters the degraded mode, the concurrent throughput of the disk array decreases, and the average response time of read and write I / O increases. Therefore, how to rebuild data to the hot spare disk in the shortest possible time after a disk in the disk array fails is a very important problem that needs to be solved urgently. Especially when the disk capacity increases rapidly but the disk speed does not increase, it takes a very long time to rebuild the ...

AI Technical Summary

Problems solved by technology

[0016] The existing disk array reconstruction process has the following problems: In the scenario of mixed read and write loads, the head of the hot spare disk will be between "rebuild data area", "rebuild data area by reading" and "rebuild data area by writing" Moving, frequent head seek operations will greatly reduce the reconstruction bandwidth of the hot spare disk, thereby prolonging the reconstruction time of the disk array and increasing the possibility of data loss

Images