System and method for disk sector failure prediction

• US 8,760,780 B1
• Filed: 03/22/2012
• Issued: 06/24/2014
• Est. Priority Date: 03/22/2012
• Status: Expired due to Fees

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Brooker, Marc J.
• Primary Examiner(s): Dinh, Tan X
• Application Number:
  US13/427,788
• Time in Patent Office: 824 Days
• Field of Search: 360/31, 360/40, 360/47, 360/53, 360/54, 711/712, 714/794, 714/774
• US Class Current: 360/31
• CPC Class Codes: G11B 20/1816 Testing
• CPC Class Codes: G11B 2220/2508 Magnetic discs
• CPC Class Codes: G11B 2220/2516 Hard disks
• CPC Class Codes: G11B 27/36 Monitoring, i.e. supervising the progress of recording or reproducing for digital recording G11B20/00 and s.gr.; for monitoring, testing or measuring of TV recorders of the type covered by H04N5/76 and subgroups, see H04N17/06

Abstract

A system and method for predicting the likelihood of failure of the individual sectors of a magnetic storage disk based upon the monitoring of adjacent sector performance in responding to access requests. The prediction for a specific target sector'"'"'s fidelity can be made through the gathering of sector performance data that occurs during normal read/write actions to the adjacent sectors, without a recent access request necessarily being made to the target sector. Scrubbing of the sectors can also be directed based upon the needed sector access data for target sector fidelity prediction.

Claims

1. A system for predicting sector failure on a magnetic storage disk, comprising:

at least one magnetic storage disk having a plurality of sectors thereupon, the sectors selectively storing data, at least some of the sectors being adjacent to one another;

at least one head configured to selectively access data stored on the plurality of sectors on the at least one magnetic storage disk; and

at least one processor and a memory storing computer-readable instructions that, when executed by the at least one processor, causes the system to at least;

access data for data accesses by the head at individual sectors within the plurality of sectors; and

determine a likelihood of failure of a first sector of the plurality of sectors without accessing data for data accesses by the head at the first sector, the likelihood determined based at least in part upon access of the head to a second sector within the plurality of sectors and a predicted relationship between failures in the first and second sectors and relative locations of the first and second sectors on the at least one magnetic storage disk.

2. The system of claim 1, wherein the head selectively accesses the data at each sector through a read or write operation.

3. The system of claim 1, wherein the first sector and second sector are contiguous.

4. The system of claim 1, wherein the first sector and second sector are proximal to each other and non-contiguous.

5. The system of claim 1, wherein the likelihood of the failure of the first sector is based upon successful access of the head to a plurality of sectors adjacent to the first sector within the plurality of sectors.

6. A system for predicting, at least, sector failure on a magnetic storage disk, comprising:

at least one magnetic storage disk having a plurality of sectors thereupon, the sectors selectively storing data, at least some of the sectors being adjacent one another, wherein the sectors are selectively accessible; and

at least one processor and a memory storing computer-readable instructions that, when executed by the at least one processor, causes the system to at least;

access data for data accesses at individual sectors within the plurality of sectors and

determine a likelihood of failure of a first sector of the plurality of sectors without accessing data for data accesses by the head at the first sector, the likelihood determined based at least in upon one or more accesses to a second sector within the plurality of sectors and a predicted relationship between failures in the first and second sectors and relative locations of the first and second sectors on the at least one magnetic storage disk.

7. The system of claim 6, wherein the one or more accesses is through a read or write operation.

8. The system of claim 6, wherein the first sector and second sector are contiguous.

9. The system of claim 6, wherein the first sector and second sector are proximal to each other and non-contiguous.

10. The system of claim 6, wherein the likelihood of the failure of the first sector is based upon access to a plurality of adjacent sectors within the plurality of sectors.

11. The system of claim 6, wherein the likelihood of the failure of the first sector is based upon intentional access to at least one adjacent second sector within the plurality of sectors.

12. The system of claim 6, wherein the likelihood of the failure of the first sector is based upon one or more accesses to the second sector within a predetermined time period.

13. The system of claim 6, wherein at least some of the plurality of sectors are scrubbed to determine sector failures, and the likelihood of the failure of the first sector is based upon one or more accesses to at least one adjacent scrubbed sector.

14. The system of claim 13, wherein the scrubbing is done to desired sectors based upon a predetermined pattern to collect sufficient data to predict the likelihood of the failure.

15. A method for predicting, at least, sector failure on a magnetic storage disk, comprising:

selectively accessing at least one sector of a magnetic storage disk having a plurality of sectors thereupon, the magnetic storage disk configured to selectively store data on the plurality of sectors, and at least seem of the sectors being adjacent to one another;

receiving data for data accesses at individual sectors within the plurality of sectors; and

determining a likelihood of failure of a first sector based at least in part upon one or more data accesses to at least one adjacent sector of the plurality of sectors without accessing data for data accesses by the head at the first sector, the likelihood determined based at least in part on a predicted relationship between failures in the first sector and the at least one adjacent sector and relative locations of the first sector and the at least one adjacent sector on the magnetic storage disk.

16. The method of claim 15, wherein the data accesses comprise accessing data through a read or write operation.

17. The method of claim 15, wherein the at least one adjacent sector is contiguous to the first sector.

18. The method of claim 15, wherein the first sector and the at least one adjacent sector are proximal to each other and non-contiguous.

19. The method of claim 15, wherein the likelihood of failure is based upon the access to a plurality of adjacent sectors within the plurality of sectors.

20. The method of claim 15, wherein the likelihood of failure is based upon one or more intentional accesses to the at least one adjacent sector.

21. The method of claim 15, wherein the likelihood of failure is based upon one or more accesses to the at least one adjacent sector within a predetermined time period.

22. The method of claim 15, further comprising scrubbing at least some of the plurality of sectors to identify failed sectors, and wherein the likelihood of failure of the first sector is based upon one or more accesses to at least one adjacent scrubbed sector.

23. The method of claim 22, wherein said scrubbing is performed on sectors based upon a predetermined pattern.

24. A non-transitory computer-readable storage medium containing instructions that, when executed by one or more processors of a system, cause the system to at least:

receive data for data accesses at individual sectors of a magnetic storage disk having a plurality of sectors thereupon, at least some of the sectors being adjacent to other sectors within the plurality of sectors; and

determining a likelihood of failure of a first sector based upon one or more accesses to at least one sector adjacent to the first sector, wherein the likelihood of failure is based at least in part on a predicted relationship between failures in the first sector and the adjacent sector and relative locations of the first sector and the adjacent sector.

25. The storage medium of claim 24, wherein the data for data accesses comprises read or write operation data for individual sectors.

26. The storage medium of claim 24, wherein the adjacent sector is contiguous with the first sector.

27. The storage medium of claim 24, wherein the the first sector and adjacent sector are proximal to each other and non-contiguous.

28. The storage medium of claim 24, wherein the likelihood of failure is based upon access to data pertaining to a plurality of adjacent sectors within the plurality of sectors.

29. The storage medium of claim 24, wherein the likelihood of failure is based upon one or more intentional accesses to the adjacent sector.

30. The storage medium of claim 24, wherein the likelihood of failure is based upon one or more accesses to the adjacent sector within a predetermined time period.

31. The storage medium of claim 24, further comprising instructions that, when executed by the one or more processors, at least cause the system to scrub at least some of the plurality of sectors to identify failed sectors, and wherein the likelihood of failure is based upon one or more accesses to at least one adjacent scrubbed sector.

32. The storage medium of claim 31, further comprising instructions that, when executed by the one or more processors, at least cause the system to generate a predetermined pattern of scrubbed sectors.