RAID system using regional error statistics for redundancy grouping

• US 20060224914A1
• Filed: 03/31/2005
• Published: 10/05/2006
• Est. Priority Date: 03/31/2005
• Status: Active Grant

• Current Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Original Assignee: Hitachi Global Storage Technologies Netherlands BV (Western Digital Corporation)
• Inventors: Blaum, Mario, Hoyt, Roger
• Granted Patent: US 7,484,137 B2
• Time in Patent Office: Days
• US Class Current: 714/6
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 2211/109 Sector level checksum or ECC, i.e. sector or stripe level checksum or ECC in addition to the RAID parity calculation

Abstract

Geometrically-dependent error rates are used to identify sectors for XORing data in a RAID system for parity purposes in such a way that the probability of failure of any particular group is minimized.

Claims

1. RAID-implemented software embodying logic comprising:

associating at least a first error statistic with a first geometric disk location;

associating at least a second error statistic with a second geometric disk location; and

rendering, based on the error statistics, at least one parity output by combining data on a first disk in the first geometric disk location with data on a second disk in the second geometric disk location to minimize the probability of failure of a data group.

2. The RAID-implemented software of claim 1, wherein the logic further comprises establishing a redundant region on a redundant disk for storing the parity output based at least in part on geometry-dependent error statistics.

3. The RAID-implemented software of claim 2, wherein the redundant region is in a different radial location than the first and second disk locations.

4. The RAID-implemented software of claim 1, wherein the error statistics are static.

5. The RAID-implemented software of claim 1, wherein the error statistics are adaptively determined using error correction code (ECC) information associated with the RAID.

7. The RAID controller of claim 5, wherein the RAID operation includes combining data in a first region on a first disk at least with data in a second region on a second disk to render a parity output for storage in at least one region of at least one redundant disk.

8. The RAID controller of claim 7, wherein the first region is in a first geometric disk location and the second region is in a second geometric disk location different than the first geometric disk location, the regions being established at least in part based on the first region being associated with a relatively better error rate than an error rate associated with the second region.

9. The RAID controller of claim 5, wherein the RAID operation includes selecting a region on a redundant disk for storing a parity output generated by combining data in a first region on a first disk at least with data in a second region on a second disk, and the region on the redundant disk is in a geometric disk location that is different at least from the geometric disk location of the first or second region.

10. The RAID controller of claim 9, wherein the RAID operation further includes combining data in the first region on the first disk at least with data in the second region on the second disk to render the parity output, wherein the first region is in a first geometric disk location and the second region is in a second geometric disk location different than the first geometric disk location, the first and second regions being established at least in part based on the first region being associated with a relatively better error rate than an error rate associated with the second region.

11. The RAID controller of claim 5, wherein the error information includes error statistics and the region is a sector.

6. A RAID controller executing logic to store data in plural storage disks, the logic comprising:

receiving at least two error informations respectively associated with two geometric disk regions, the error informations at least indicating different error rates from each other; and

using the error informations to execute at least one RAID operation.

12. The RAID controller of claim 6, wherein the error information is static.

13. The RAID controller of claim 6, wherein the error information is adaptively determined using error correction code (ECC) information associated with the RAID.

14. A disk drive controller, comprising:

means for associating a first error statistic with a first disk region and a second error statistic with a second disk region;

means for combining data from a first region on a first disk at least with data from a second region on a second disk; and

means for identifying the first and second regions using the first and second error statistics.

15. The disk drive controller of claim 14, wherein the controller is implemented in a RAID system.

16. The disk drive controller of claim 14, wherein the means for identifying minimizes the probability of failure of a data group based on the error statistics.

17. The disk drive controller of claim 14, further comprising means for establishing a redundant region on a redundant disk for storing the result from the means for combining based at least in part on geometry-dependent error statistics.

18. The disk drive controller of claim 17, wherein the redundant region is in a different radial location than the first and second regions.

19. The disk drive controller of claim 14, wherein the error statistics are static.

20. The disk drive controller of claim 14, comprising means for adaptively determining error statistics are using error correction code (ECC) information associated with the RAID.