Systems and methods for sequential resilvering

• US 9,594,635 B2
• Filed: 09/04/2014
• Issued: 03/14/2017
• Est. Priority Date: 06/09/2014
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Taylor, Eric Carl, Sanouillet, Blaise, Bourbonnais, Roch
• Primary Examiner(s): Leibovich, Yair
• Application Number:
  US14/477,645
• Publication Number:
  US 20150355968A1
• Time in Patent Office: 922 Days
• Field of Search: 714/6.2
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1092 Rebuilding, e.g. when physically replacing a failing disk

Abstract

Implementations claimed and described herein provide systems and methods for the efficient rebuilding of a failed storage device through sequential resilvering. In one implementation, blocks for resilvering are discovered. The blocks correspond to input/output requests not successfully completed for a failed storage device. A coarse grained sorting of the blocks is performed based on a block location of each of the blocks on the failed storage device. The block locations of the blocks are stored in memory according to the coarse grained sorting. A fine grained sorting of the blocks is performed based on the coarse grained sorting of the blocks. The blocks are sequentially resilvered based on the fine grained sorting.

Claims

1. A method for reconstructing data comprising:

discovering a plurality of blocks for resilvering using at least one computing device, the blocks corresponding to input/output requests not successfully completed for a failed storage device;

coarse grained sorting the plurality of blocks using the at least one computing device, the coarse grained sorting based on a block location of each of the plurality of blocks on the failed storage device;

storing the block location of each of the plurality of blocks on disk according to the coarse grained sorting;

fine grained sorting the coarse grained sorted plurality of blocks further based on the block location using the at least one computing device; and

resilvering the fine grained sorted plurality of blocks sequentially according to the block location using the at least one computing device.

2. The method of claim 1, wherein the coarse grained sorting includes sorting each of the plurality of blocks into one of a plurality of temporary objects in an object array based on the block location of each of the plurality of blocks on the failed storage device.

3. The method of claim 2, wherein a number of the plurality of temporary objects in the object array depends on an amount of disk memory.

4. The method of claim 2, wherein the fine grained sorting includes reading the plurality of blocks into a sorted tree by block location for each of the plurality of temporary objects in the object array.

5. The method of claim 4, wherein the plurality of blocks are sequentially resilvered based on the sorted tree for each of the plurality of temporary objects in the object array.

6. The method of claim 1, wherein the fine grained sorting includes reading each of the plurality of blocks into a sorted tree by the block location.

7. The method of claim 6, wherein the sorted tree is an Adelson Velskii Landis tree.

8. The method of claim 1, wherein the sequential resilvering includes sequentially traversing the plurality of blocks in logical block order to reconstruct data for which the input/output requests were not successfully completed for the failed storage device.

9. One or more tangible non-transitory computer-readable storage media storing computer-executable instructions for performing a computer process on a computing system, the computer process comprising:

discovering a plurality of blocks for resilvering;

storing an object array having a first temporary object on disk;

sorting a first subset of the plurality of blocks into the first temporary object based on a block location of each of the plurality of blocks;

reading each of the plurality of blocks in the first temporary object into a sorted tree;

resilvering each of the plurality of blocks in the first temporary object sequentially based on the sorted tree; and

deleting the first temporary object following the sequential resilvering of the plurality of blocks in the first subset.

10. The one or more tangible non-transitory computer-readable storage media of claim 9, wherein the object array further includes a second temporary object.

11. The one or more tangible non-transitory computer-readable storage media of claim 10, further comprising:

sorting a second subset of the plurality of blocks into the second temporary object based on the block location of each of the plurality of blocks.

12. The one or more tangible non-transitory computer-readable storage media of claim 11, further comprising:

reading each of the plurality of blocks in the second temporary object into a second sorted tree.

13. The one or more tangible non-transitory computer-readable storage media of claim 12, further comprising:

resilvering each of the plurality of blocks in the second temporary object sequentially based on the second sorted tree after the sequential resilvering of each of the plurality of blocks in the first temporary object is completed.

14. The one or more tangible non-transitory computer-readable storage media of claim 13, further comprising:

deleting the second temporary object following the sequential resilvering of the plurality of blocks in the second subset.

15. The one or more tangible non-transitory computer-readable storage media of claim 9, wherein the sorted tree is an Adelson Velskii Landis tree.

16. A system for reconstructing data comprising:

a storage pool having a plurality of storage devices configured to store data; and

a file system in communication with the storage pool and having at least one processing unit configured to identify a failed device in the plurality of storage devices, the failed device corresponding to plurality of blocks for which input/output requests were not successfully completed, and to sequentially resilver the plurality of blocks based on a plurality of sorted trees, each of the plurality of sorted trees corresponding to one of a plurality of temporary objects in an object array.

17. The system of claim 16, wherein each of the plurality of blocks are sorted into one of the plurality of temporary objects based on a block location on the failed storage device.

18. The system of claim 16, wherein the plurality of blocks are sorted into the plurality of sorted trees by block location.

19. The system of claim 16, wherein the plurality of temporary objects in the object array are stored on disk.

20. The system of claim 16, wherein each of the plurality of sorted trees is an Adelson Velskii Landis tree.