Resilience of data storage systems by managing partial failures of solid state drives

Patent metadata

Field	Value
Publication number	US-10719398-B1
Application number	US-201715653341-A
Country	US
Kind code	B1
Filing date	Jul 18, 2017
Priority date	Jul 18, 2017
Publication date	Jul 21, 2020
Grant date	Jul 21, 2020

Abstract

Each SSD in an array is divided into splits that are contiguous LBA ranges of equal capacity. The splits are managed as if they were sub-drives or logical drives and used in RAID groups or other forms of redundancy schemes. A processor external to the SSD collects pending defect data log from the SSDs and determines which splits are associated with reported defects. Failed splits are rebuilt in spare splits using RAID or another redundancy scheme. Repaired splits are returned to service as spare splits. Irreparable splits are marked as lost. In the case of SSD failure the data splits of the SSD are rebuilt to a spare SSD. If enough failed splits of the failed SSD can be repaired then the repaired SSD may be returned to service.

First claim

What is claimed is:

1. An apparatus comprising:
a storage array comprising:
at least one computing node comprising at least one processor and non-transitory memory;
a drive array comprising a plurality of SSDs (solid state drives), the drive array in communication with the computing node; and
at least one processor, external with respect to the SSDs, that implements instructions comprising:
organizing instructions that organize the SSDs as a plurality of splits, each split comprising a contiguous range of logical addresses presented by one of the SSDs and having a capacity equal to other splits of the plurality of splits;
grouping instructions that group a plurality of spare splits of the plurality of splits into a redundancy group;
monitoring instructions that detect a failed split of the redundancy group on a non-failed one of the SSDs; and
rebuilding instructions that rebuild the failed split on a spare split.

2. The apparatus of claim 1 wherein the organizing instructions organize the splits into logical slices of equal capacity corresponding to a track, a cylinder, a fraction of SSD capacity, or a percentage of SSD capacity.

3. The apparatus of claim 1 wherein the logical addresses are LBAs (logical block addresses).

4. The apparatus of claim 1 wherein non-contiguous physical addresses map to the contiguous range of logical addresses of each split.

5. The apparatus of claim 4 wherein the non-contiguous physical addresses of each split map to a plurality of dies.

6. The apparatus of claim 1 wherein the monitoring instructions process pending defect data logs from the SSDs to identify splits with which reported defects are associated.

7. The apparatus of claim 6 wherein the rebuilding instructions select the spare split from an SSD having a greatest number of spare splits.

8. The apparatus of claim 1 comprising remediation instructions that attempt to repair the failed split.

9. The apparatus of claim 8 wherein the remediation instructions return repaired splits to service as spare splits and mark unrepairable splits as lost splits.

10. The apparatus of claim 1 wherein the rebuilding instructions rebuild all failed data splits of a failed SSD on a spare SSD, and the remediation instructions return repaired SSDs to service and mark unrepairable SSDs as lost.

11. A method comprising:
in a storage array comprising at least one computing node comprising at least one processor and non-transitory memory, and a drive array comprising a plurality of SSDs (solid state drives), the drive array in communication with the computing node:
externally with respect to the SSDs, organizing the SSDs as a plurality of splits, each split comprising a contiguous range of logical addresses presented by one of the SSDs and having a capacity equal to other splits of the plurality of splits;
grouping a plurality of spare splits of the plurality of splits into a redundancy group;
detecting a failed split of the redundancy group on a non-failed one of the SSDs; and
rebuilding the failed split on a spare split.

12. The method of claim 11 comprising organizing the splits into logical slices of equal capacity corresponding to a track, a cylinder, a fraction of SSD capacity, or a percentage of SSD capacity.

13. The method of claim 11 wherein the logical addresses are LBAs (logical block addresses), and non-contiguous physical addresses map to the contiguous range of logical addresses of each split, and the non-contiguous physical addresses of each split map to a plurality of dies, and comprising organizing each split as a contiguous range of LBAs.

14. The method of claim 11 comprising processing pending defect data logs from the SSDs to identify splits with which reported defects are associated.

15. The method of claim 14 comprising selecting the spare split from an SSD having a greatest number of spare splits.

16. The method of claim 11 comprising attempting to repair the failed split.

17. The method of claim 16 comprising returning repaired splits to service as spare splits.

18. The method of claim 17 comprising marking unrepairable splits as lost splits.

19. The method of claim 11 comprising rebuilding all failed data splits of a failed SSD on a spare SSD.

20. The method of claim 19 comprising returning repaired SSDs to service and marking unrepairable SSDs as lost.

Assignees

• Emc Ip Holding Co Llc

Inventors

• Li Jun
• Guyer James Marriot
• Garvey Michael D
• Ives Stephen Richard

Classifications

• G06F11/1088

  Reconstruction on already foreseen single or plurality of spare disks · CPC title

• G06F3/0644

  Management of space entities, e.g. partitions, extents, pools · CPC title

• G06F3/0619

  in relation to data integrity, e.g. data losses, bit errors · CPC title

• G06F3/0616

  in relation to life time, e.g. increasing Mean Time Between Failures [MTBF] · CPC title

• G06F3/0688

  Non-volatile semiconductor memory arrays · CPC title