Re-allocation of disks based on disk health prior to restore
Disk drive reallocation or replacement is disclosed. When performing a data protection operation, a health score is determined for each of the disk drives associated with the data protection operation. Replacement is recommended for each of the disk drives with an unfavorable health score. The recommendation may also identify a drive class based in part on the write or wear pattern. This allows unhealthy drives to be replaced prior to performing the data protection operation.
1. A method comprising:
initiating a data protection operation associated with target disk drives of a storage system that includes a set of disk drives, the target disk drives included in the set of disk drives;
determining a failure probability for each of the disk drives of the storage system that is associated with the data protection operation using a failure prediction model;
generating a drive replacement recommendation for each of the disk drives based on the failure probability and a write pattern prior to the data protection operation, wherein the drive replacement recommendation includes a drive class associated with each of the disk drives, wherein the failure probability score represents a health score, wherein the drive class is based on the write pattern;
replacing the disk drives recommended for replacement, wherein drive classes of the replaced disk drives are selected to handle the write patterns;
reallocating disk drives of the storage system for the data protection operation such that new target disk drives used by the data protection operation only include disk drives from the set of disk drives with a sufficient health score;
changing a mapping of virtual drives of the data protection operation such that the virtual drives are mapped to the new target disk drives; and
performing the data protection operation to the new target disk drives whose failure probability satisfies a threshold failure probability.
2. The method of claim 1 , wherein the data protection operation comprises a restore operation configured to restore data to the new target disk drives, wherein the drive replacement recommendation when the data protection operation is a restore operation differs from the drive replacement recommendation when the data protection operation is a backup operation.
3. The method of claim 1 , further comprising retrieving attributes of each of the disk drives of the storage system.
4. The method of claim 3 , further comprising training the failure prediction model using historical attributes collected from the disk drives.
5. The method of claim 1 , further comprising replacing the disk drives recommended for replacement with new disk drives.
6. The method of claim 5 , wherein the drive class is based at least on a write pattern associated with the disk drives.
7. The method of claim 1 , further comprising generating a health score for each of the disk drives based on corresponding failure probabilities and recommending replacement of the disk drives whose health scores to not satisfy a threshold health score.
8. The method of claim 1 , wherein the data protection operation comprises a backup operation or a restore operation.
9. The method of claim 8 , wherein the data protection operation is performed after the drive replacement recommendation is made, reallocation of the disk drives is performed, and the changing of the mapping of virtual drives to physical drives is performed.
10. The method of claim 1 , wherein the health score is an expected time to failure or a time during which drive failure is expected.
11. A non-transitory storage medium having stored therein instructions that are executable by one or more hardware processors to perform operations comprising:
initiating a data protection operation associated with target disk drives of a storage system that includes a set of disk drives, the target disk drives included in the set of disk drives;
determining a failure probability for each of the disk drives of the storage system that is associated with the data protection operation using a failure prediction model;
generating a drive replacement recommendation for each of the disk drives based on the failure probability and a write pattern prior to the data protection operation, wherein the drive replacement recommendation includes a drive class associated with each of the disk drives, wherein the failure probability score represents a health score, wherein the drive class is based on the write pattern;
replacing the disk drives recommended for replacement, wherein drive classes of the replaced disk drives are selected to handle the write patterns;
reallocating disk drives of the storage system for the data protection operation such that new target disk drives used by the data protection operation only include disk drives from the set of disk drives with a sufficient health score;
changing a mapping of virtual drives of the data protection operation such that the virtual drives are mapped to the new target disk drives; and
performing the data protection operation to the new target disk drives whose failure probability satisfies a threshold failure probability.
12. The non-transitory storage medium of claim 11 , wherein the data protection operation comprises a restore operation configured to restore data to the new target disk drives, wherein the drive replacement recommendation when the data protection operation is a restore operation differs from the drive replacement recommendation when the data protection operation is a backup operation.
13. The non-transitory storage medium of claim 11 , further comprising retrieving attributes of each of the disk drives.
14. The non-transitory storage medium of claim 13 , further comprising training the failure prediction model using historical attributes collected from the disk drives.
15. The non-transitory storage medium of claim 11 , further comprising replacing the disk drives recommended for replacement with new disk drives.
16. The non-transitory storage medium of claim 15 , wherein the drive class is based at least on a write pattern associated with the disk drives.
17. The non-transitory storage medium of claim 11 , further comprising generating a health score for each of the disk drives based on corresponding failure probabilities and recommending replacement of the disk drives whose health scores to not satisfy a threshold health score.
18. The non-transitory storage medium of claim 11 , wherein the data protection operation comprises a backup operation or a restore operation.
19. The non-transitory storage medium of claim 18 , wherein the data protection operation is performed after the drive replacement recommendation is made, reallocation of the disk drives is performed, and the changing of the mapping of virtual drives to physical drives is performed.
20. The non-transitory storage medium of claim 19 , wherein the health score is an expected time to failure or a time during which drive failure is expected.