Storage system with prioritized RAID rebuild

1. An apparatus comprising:

a storage system comprising a plurality of storage devices;

the storage system being configured:

to establish a redundant array of independent disks (RAID) arrangement comprising a plurality of stripes each having multiple portions distributed across multiple ones of the storage devices;

to detect a failure of at least one of the storage devices;

responsive to the detected failure, to determine for each of two or more remaining ones of the storage devices a number of stripe portions, stored on that storage device, that are part of stripes impacted by the detected failure; and

to prioritize a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions;

wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

prioritizing a first one of the remaining storage devices having a relatively low determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes, over a second one of the remaining storage devices having a relatively high determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes.

2. The apparatus of claim 1 wherein the RAID arrangement supports at least one recovery option for reconstructing data blocks of at least one of the storage devices responsive to a failure of that storage device.

3. The apparatus of claim 2 wherein the RAID arrangement comprises a RAID 6 arrangement supporting recovery from failure of up to two of the storage devices.

4. The apparatus of claim 1 wherein the stripe portions of each of the stripes comprise a plurality of data blocks and one or more parity blocks.

5. The apparatus of claim 1 wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

selecting for rebuilding of its stripe portions that are part of the impacted stripes the particular one of the remaining storage devices that has a lowest determined number of stripe portions relative to the determined numbers of stripe portions of the one or more other remaining storage devices.

6. The apparatus of claim 1 wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

determining health measures for respective ones of the remaining storage devices; and

taking the determined health measures into account in selecting the particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes.

7. The apparatus of claim 1 wherein the storage system is further configured to balance the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices.

8. The apparatus of claim 7 wherein balancing the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices comprises:

maintaining rebuild work statistics for each of the remaining storage devices over a plurality of iterations of a rebuild process for rebuilding the stripe portions of the impacted stripes; and

selecting different subsets of the remaining storage devices to participate in respective different iterations of the rebuild process based at least in part on the rebuild work statistics.

9. The apparatus of claim 8 wherein maintaining rebuild work statistics comprises maintaining a work counter vector that stores counts of respective rebuild work instances for respective ones of the remaining storage devices and wherein a decay factor is applied to the work counter vector in conjunction with one or more of the iterations.

10. The apparatus of claim 7 wherein balancing the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices comprises:

tracking amounts of rebuild work performed by respective ones of the remaining storage devices in rebuilding the stripe portions of a first one of the impacted stripes; and

excluding at least one of the remaining storage devices from performance of rebuild work for another one of the impacted stripes based at least in part on the tracked amounts of rebuild work for the first impacted stripe;

wherein said at least one excluded remaining storage device for the other one of the impacted stripes comprises the remaining storage device that performed a largest amount of rebuild work of the amounts of rebuild work performed by respective ones of the remaining storage devices for the first impacted stripe.

11. An apparatus comprising:

a storage system comprising a plurality of storage devices;

the storage system being configured:

to establish a redundant array of independent disks (RAID) arrangement comprising a plurality of stripes each having multiple portions distributed across multiple ones of the storage devices;

to detect a failure of at least one of the storage devices;

responsive to the detected failure, to determine for each of two or more remaining ones of the storage devices a number of stripe portions, stored on that storage device, that are part of stripes impacted by the detected failure; and

to prioritize a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions;

wherein determining for one of the remaining storage devices the number of stripe portions, stored on that storage device, that are part of the impacted stripes comprises:

determining a number of data blocks stored on that storage device that are part of the impacted stripes;

determining a number of parity blocks stored on that storage device that are part of the impacted stripes; and

summing the determined number of data blocks and the determined number of parity blocks to obtain the determined number of stripe portions for that storage device.

12. An apparatus comprising:

a storage system comprising a plurality of storage devices;

the storage system being configured:

to establish a redundant array of independent disks (RAID) arrangement comprising a plurality of stripes each having multiple portions distributed across multiple ones of the storage devices;

to detect a failure of at least one of the storage devices;

responsive to the detected failure, to determine for each of two or more remaining ones of the storage devices a number of stripe portions, stored on that storage device, that are part of stripes impacted by the detected failure; and

to prioritize a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions;

wherein the storage system is further configured:

to rebuild, for the particular prioritized one of the remaining storage devices, its stripe portions that are part of the impacted stripes;

to select another one of the remaining storage devices for rebuild prioritization; and

to rebuild, for the selected other one of the remaining storage devices, its stripe portions that are part of the impacted stripes.

13. The apparatus of claim 12 wherein the selecting of another one of the remaining storage devices for rebuild prioritization and the rebuilding, for the selected other one of the remaining storage devices, its stripe portions that are part of the impacted stripes, are repeated for one or more additional ones of the remaining storage devices until all of the stripe portions of the impacted stripes are fully rebuilt.

14. A method for use in a storage system comprising a plurality of storage devices, the method comprising:

establishing a redundant array of independent disks (RAID) arrangement comprising a plurality of stripes each having multiple portions distributed across multiple ones of the storage devices;

detecting a failure of at least one of the storage devices;

responsive to the detected failure, determining for each of two or more remaining ones of the storage devices a number of stripe portions, stored on that storage device, that are part of stripes impacted by the detected failure; and

prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions;

wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

prioritizing a first one of the remaining storage devices having a relatively low determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes, over a second one of the remaining storage devices having a relatively high determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes.

15. The method of claim 14 wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

selecting for rebuilding of its stripe portions that are part of the impacted stripes the particular one of the remaining storage devices that has a lowest determined number of stripe portions relative to the determined numbers of stripe portions of the one or more other remaining storage devices.

16. The method of claim 14 further comprising balancing the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices.

17. A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, wherein the program code when executed by a processor of a storage system comprising a plurality of storage devices, causes the storage system:

to establish a redundant array of independent disks (RAID) arrangement comprising a plurality of stripes each having multiple portions distributed across multiple ones of the storage devices;

to detect a failure of at least one of the storage devices;

responsive to the detected failure, to determine for each of two or more remaining ones of the storage devices a number of stripe portions, stored on that storage device, that are part of stripes impacted by the detected failure; and

to prioritize a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions;

wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

prioritizing a first one of the remaining storage devices having a relatively low determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes, over a second one of the remaining storage devices having a relatively high determined number of stripe portions for rebuilding of its stripe portions that are part of the impacted stripes.

18. The computer program product of claim 17 wherein prioritizing a particular one of the remaining storage devices for rebuilding of its stripe portions that are part of the impacted stripes, based at least in part on the determined numbers of stripe portions, comprises:

selecting for rebuilding of its stripe portions that are part of the impacted stripes the particular one of the remaining storage devices that has a lowest determined number of stripe portions relative to the determined numbers of stripe portions of the one or more other remaining storage devices.

19. The computer program product of claim 17 wherein the program code when executed by the processor of the storage system further causes the storage system to balance the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices.

20. The computer program product of claim 17 wherein balancing the rebuilding of the stripe portions of the impacted stripes across the remaining storage devices comprises:

maintaining rebuild work statistics for each of the remaining storage devices over a plurality of iterations of a rebuild process for rebuilding the stripe portions of the impacted stripes; and

selecting different subsets of the remaining storage devices to participate in respective different iterations of the rebuild process based at least in part on the rebuild work statistics.