Method and system for increasing parallelism of disk accesses when restoring data in a disk array system

Abstract

In a disk array environment such as a RAID-6 environment, the overall performance overhead associated with exposed mode operations such as resynchronization, rebuild and exposed mode read operations is reduced through increased parallelism. By selecting only subsets of the possible disks required to solve a parity stripe equation for a particular parity stripe, accesses to one or more disks in a disk array may be omitted, thus freeing the omitted disks to perform other disk accesses. In addition, disk accesses associated with different parity stripes may be overlapped such that the retrieval of data necessary for restoring data for one parity stripe is performed concurrently with the storage of restored data for another parity stripe.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application is related to the following U.S. patent applications all filed on even date herewith by Carl Edward Forhan, Robert Edward Galbraith and Adrian Cuenin Gerhard: Ser. No. ______, entitled “METHOD AND SYSTEM FOR ENHANCED ERROR IDENTIFICATION WITH DISK ARRAY PARITY CHECKING,” Ser. No. ______, entitled “RAID ENVIRONMENT INCORPORATING HARDWARE-BASED FINITE FIELD MULTIPLIER FOR ON-THE-FLY XOR,” Ser. No. ______, entitled “METHOD AND SYSTEM FOR IMPROVED BUFFER UTILIZATION FOR DISK ARRAY PARITY UPDATES,” and Ser. No. ______, entitled “METHOD AND SYSTEM FOR RECOVERING FROM ABNORMAL INTERRUPTION OF A PARITY UPDATE OPERATION IN A DISK ARRAY SYSTEM.” Each of these applications is incorporated by reference herein.FIELD OF THE INVENTION [0002] The present invention relates to data protection methods for data storage and, more particularly, to systems implementing RAID-6 and similar data protection and recovery strategies. BACKGROUND OF ...

AI Technical Summary

Benefits of technology

[0018] In another aspect, parallelism may be increased in a disk array system by overlapping disk accesses associated with different parity stripes when restoring data in a disk array (e.g., to resynchronize parity and data, or to rebuild data for an exposed disk). Specifically, consistent with this aspect of the invention, restoring data to a disk array may include the retrieval of a first set of data associated with a first parity stripe, coupled with concurrent operations of writing to the disk array a result value generated by processing the first set of data, and reading from the disk array a second set of data associated with a second parity stripe. By overlapping read and write accesses associated with different parity stripes, data associated with multiple parity stripes may be restored with less overhead than if the accesses and operations associated with restoring data to different parity stripes were performed sequentially.

Problems solved by technology

In the former instance, the RAID algorithms are packaged into separate controller hardware coupled to the computer input / output (“I / O”) bus and, although adding little or no central processing unit (“CPU”) overhead, the additional hardware required nevertheless adds to the overall system cost.

On the other hand, software implementations incorporate the RAID algorithms into system software executed by the main processor together with the operating system, obviating the need and cost of a separate hardware controller, yet adding to CPU overhead.

The mathematics of implementing RAID-6 involve complicated calculations that are also repetitive.

For example, one limitation of existing RAID-6 designs relates to the performance overhead associated with performing resync (where parity data for a parity stripe is resynchronized with the current data), rebuild (where data from a faulty or missing drive is regenerated based upon the parity data) or other exposed mode operations such as exposed mode reads.

In each of these exposed mode operations, however, the requirements of reading data from certain disks and writing data back to certain disks results in substantial performance overhead, specifically with respect to the sequential nature of the various disk access operations on the disk array.

Images