System and method for reducing unrecoverable media errors in a disk subsystem

• US 7,447,938 B1
• Filed: 05/03/2007
• Issued: 11/04/2008
• Est. Priority Date: 04/24/2003
• Status: Expired due to Term

• Current Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Original Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Inventors: Coatney, Douglas W.
• Primary Examiner(s): Lohn; Joshua A
• Application Number:
  US11/799,795
• Time in Patent Office: 551 Days
• Field of Search: 714/5, 714/6, 714/7, 714/8, 714/42, 714/770
• US Class Current: 714/6.32
• CPC Class Codes: G06F 11/1084 Degraded mode, e.g. caused by single or multiple storage removals or disk failures

Abstract

A system and method for reducing unrecoverable errors in a disk drive by undertaking an aggressive reassignment of slow-reading or currently recoverable-but-erroneous sectors to the spares pool is provided. A recovered error is treated by the system operating system as a fatal error and thereby the sectors involved are reassigned to the spares pool immediately. Reassignment is recommended by a reassignment utility at the disk interface level, which passes a status up to the RAID subsystem by which reassignment is performed. In order to prevent a double-disk panic, the RAID subsystem is instructed to ignore reassignment recommendations of this type (e.g. reassignment of recoverable errors) if the RAID group is currently operating in a degraded state. However, if the RAID group is undegraded, then immediate assignment of the sectors encountering the recoverable error is undertaken.

Claims

1. A system for reassigning a sector encountering a recoverable error comprising:

a reassignment utility for recognizing a recoverable error on a storage device in response to an I/O operation and providing a reassignment status indicator to a storage layer included in a storage operating system; and

in the storage layer, a mechanism that determines whether the storage device is in a degraded state and adapted to reassign a sector encountering the recoverable error to a spare sector if the storage device is not in a degraded state and to store the I/O operation in a log to reassign to the spare sector at a future time if the disk is in a degraded state.

2. The system as set forth in claim 1 wherein the reassignment utility recognizes the recoverable error in response to an error-reporting signal provided by the storage device.

3. The system as set forth in claim 1 wherein the reassignment utility recognizes an error of a type including a recoverable error and a non-recoverable error on the storage device and, responsive to the reassignment utility, the mechanism in the storage layer acts (i) to reassign a sector encountering a non-recoverable error to a spare sector, and (ii) to reassign a sector encountering a recoverable error only if the storage device is not in a degraded state and to not reassign the sector if the storage device is in a degraded state.

4. The system as set forth in claim 3 wherein the spare sector is one of a plurality of spare sectors included in a spares pool on the storage device and is mapped via a defect list associated with the storage device.

5. A method for reassigning a sector encountering a recoverable error on a disk, comprising:

recognizing, with a reassignment utility, a recoverable error on a disk in response to a disk I/O operation and providing a reassignment status indicator to a disk storage layer included in a storage operating system; and

in the disk storage layer, determining whether the disk is in a degraded state and reassigning the sector encountering the recoverable error to a spare disk sector if the disk is not in a degraded state and storing the I/O operation in a log to reassign to the spare sector at a future time if the disk is in a degraded state.

6. The method as set forth in claim 5 wherein the recognizing step includes a step of receiving an error reporting signal from the disk indicative of a recoverable error condition on the disk.

7. The method as set forth in claim 5 further comprising recognizing, with the reassignment utility, an error of a type including a recoverable error and a non-recoverable error on disk and, responsive to the recognizing step, the mechanism in the disk storage layer acting (i) to reassign a disk sector encountering a non-recoverable error to a spare disk sector, and (ii) to reassign a disk sector encountering a recoverable error only if the disk is not in a degraded state and to not reassign the disk sector if the disk is in a degraded state.

8. The method as set forth in claim 7 wherein the spare disk sector is one of a plurality of spare disk sectors included in a spares pool on the disk and is mapped via a defect list associated with the disk.

9. A computer-readable medium for reassigning a sector encountering a recoverable error on a disk including program instructions for performing the steps of:

recognizing, with a reassignment utility, a sector encountering a recoverable error on a disk in response to a disk I/O operation and providing a reassignment status indicator to a disk storage layer included in a storage operating system; and

in the disk storage layer, determining whether the disk is in a degraded state and reassigning the sector encountering the recoverable error to a spare disk sector if the disk is not in a degraded state and storing the I/O operation in a log to reassign to the spare sector at a future time if the disk is in a degraded state.

10. The computer-readable medium as set forth in claim 9 wherein the recognizing step includes a step of receiving an error reporting signal from the disk indicative of a recoverable error condition on the disk.

11. The computer-readable medium as set forth in claim 10 further comprising the steps of recognizing, with the reassignment utility, an error of a type including a recoverable error and a non-recoverable error on disk and, responsive to the recognizing step, the mechanism in the disk storage layer acting (i) to reassign a disk sector encountering a non-recoverable error to a spare disk sector, and (ii) to reassign a disk sector encountering a recoverable error only if the disk is not in a degraded state and to not reassign the disk sector if the disk is in a degraded state.

12. The computer-readable medium as set forth in claim 11 wherein the spare disk sector is one of a plurality of sectors included in a spares pool on the disk and is mapped via a defect list associated with the disk.

13. A method for reducing unrecoverable errors in a disk of a RAID, group comprising:

detecting a recoverable media error in an I/O operation;

in response to the detecting, providing a recommendation to the RAID layer to reassign a sector of the disk encountering the recoverable media error to a spare sector in a pool of spare sectors on the disk of the RAID group;

reassigning, by the RAID group, the sector encountering the recoverable media error if the RAID group is in a non-degraded state; and

storing, by the RAID group, the recommendation in a log and not reassigning the recoverable media error if the RAID group is in a degraded state.

14. The method as set forth in claim 13 further comprising the steps of recognizing, with the reassignment utility, an error of a type including a recoverable error and a non-recoverable error on disk and, responsive to the recognizing step, the mechanism in the disk storage layer acting (i) to reassign a disk sector encountering a non-recoverable error to a spare disk sector, and (ii) to reassign a disk sector encountering a recoverable error only if the disk is not in a degraded state and to not reassign the disk sector if the disk is in a degraded state.

15. The method as set forth in claim 14 wherein the detecting of the recoverable media error comprises a detecting SCSI 01 status.

16. The method of claim 13, wherein the RAID group is in the degraded state after a disk failure occurs to another disk within the RAID group but before reconstruction has completed of the disk failure.

17. A storage system, including a computer system having a storage operating system and a plurality of storage devices managed by the storage operating system, the storage system comprising:

a storage adapter for receiving a first error signal from one or more of the storage devices indicative of a recoverable error in the one or more sectors of the storage device in response to an I/O operation, and a second error signal from one or more of the storage devices indicative of a non-recoverable error in one or more sectors of the storage device in response to an I/O operation;

a reassignment utility for providing a reassignment status indicator to a storage layer included in a storage operating system in response to the first error signal and the second error signal; and

a mechanism in the storage layer for selectively reassigning a sector encountering an error, including the non-recoverable error and the recoverable error, to a spare sector, reassigning the sector encountering the recoverable error to the spare sector if the storage device is not in a degraded state, and not reassigning the sector if the storage device encountering the recoverable error is in the degraded state.

18. The storage system as set forth in claim 17 wherein the recoverable error comprises an error recoverable by the storage device using internal error recovery mechanisms and the non-recoverable error comprises an error that is not recoverable by the storage device using internal error recovery mechanisms.

19. The storage system as set forth in claim 17 wherein the storage devices comprise disk drives, and the storage layer comprises a RAID layer of the storage operating system.

20. The storage system as set forth claim 17 wherein the sector is a smallest unit of storage that can be individually accessed for an I/O operation.

21. The system of claim 17, wherein the RAID group is in the degraded state after a disk failure occurs to another disk within the RAID group but before reconstruction has completed of the disk failure.

22. A method for reassigning a sector encountering an error, comprising:

receiving an error signal from one or more sectors of a storage device;

determining the error received in the error signal is recoverable;

recommending, in response to determining the error is recoverable, reassignment to a disk storage layer;

determining if a RAID subsystem is currently in a degraded state;

reassigning, if the RAID system is not in the degraded state, the recovered error to one or more spare sectors; and

storing, if the RAID system is in the degraded state, the recovered error in a log to reassign to the one or more spare sectors at a future time.

23. The method of claim 22, further comprising:

if the RAID system is in the degraded state, not reassigning the recovered error.

24. The method of claim 22, further comprising:

determining the error received is unrecoverable;

determining the RAID subsystem is currently in the degraded state;

applying, in response to the RAID system is in the degraded state, a mirror recovery technique to retrieve a copy of data that is lost in the unrecoverable error; and

reassigning the copy of the data to one or more spare sectors.

25. The method of claim 22, further comprising:

determining the RAID subsystem is not in the degraded state;

using, in response to the RAID system is not in the degraded state, a RAID parity technique to recalculate a copy of data that is lost in the unrecoverable error; and

reassigning the copy of the data to the one or more spare sectors.

26. The method of claim 22, wherein the RAID group is in the degraded state after a disk failure occurs to another disk within the RAID group but before reconstruction has completed of the disk failure.