Disk array enclosure with metadata journal

1. An apparatus comprising:

a storage system comprising a disk array enclosure, the disk array enclosure comprising:

at least one enclosure controller comprising at least one processing device coupled to a memory;

a cache comprising a metadata journal;

a plurality of data storage devices in communication with the at least one enclosure controller; and

a plurality of metadata storage devices in communication with the at least one enclosure controller, each metadata storage device being configured to store metadata corresponding to data stored on the plurality of storage devices;

wherein the at least one enclosure controller is configured:

to write a first stripe metadata page to the metadata storage devices, the first stripe metadata page corresponding to a stripe of data stored on the plurality of data storage devices of the disk array enclosure;

to determine that the write of the first stripe metadata page failed for a first metadata storage device of the plurality of metadata storage devices;

to add an entry to the metadata journal based at least in part on the determination that the write of the first stripe metadata page failed, the entry comprising an indication of the first metadata storage device and an indication of the stripe of data;

to set an indication in a data structure associated with the disk array enclosure that the first stripe metadata page for the stripe of data has not been written to the first metadata storage device; and

to set the first metadata storage device to a hiccup state based at least in part on the determination that the write of the first stripe metadata page failed for the first metadata storage device.

2. The apparatus of claim 1 wherein the at least one enclosure controller is further configured:

to determine that the write of the first stripe metadata page succeeded for a second metadata storage device of the plurality of metadata storage devices;

to determine that the metadata journal comprises an additional entry corresponding to the second metadata storage device, the additional entry comprising an indication of the second metadata storage device and an indication of the stripe of data; and

to invalidate the additional entry in the metadata journal based at least in part on the determination that the write of the first stripe metadata page succeeded for the second metadata storage device.

3. The apparatus of claim 1 wherein the memory comprises a dirty data structure associated with the first metadata storage device, the at least one enclosure controller being further configured to add an entry to the dirty data structure based at least in part on the determination that the write of the first stripe metadata page failed for the first metadata storage device, the entry added to the dirty data structure comprising a logical identifier corresponding to at least a portion of the stripe of data.

4. The apparatus of claim 3 wherein the at least one enclosure controller is further configured:

to receive a read operation comprising the logical identifier;

to determine that the dirty data structure associated with the first metadata storage device comprises the logical identifier; and

to attempt to obtain the first stripe metadata page from another of the plurality of metadata storage devices based at least in part on the determination that the dirty data structure associated with the first metadata storage device comprises the logical identifier.

5. The apparatus of claim 4 wherein:

the memory comprises a plurality of dirty data structures each associated with a corresponding metadata storage device of the plurality of metadata storage devices;

the at least one enclosure controller is further configured:

to maintain a copy of the first stripe metadata page on the plurality of data storage devices;

to determine that each of the dirty data structures comprises the logical identifier; and

to obtain the first stripe metadata page from the plurality of data storage devices based at least in part on the determination that the each of the dirty data structures comprises the logical identifier.

6. The apparatus of claim 4 wherein the at least one enclosure controller is further configured:

to determine that the attempt to obtain the first stripe metadata page from the another of the plurality of metadata storage devices was successful; and

to copy the first stripe metadata page to the first metadata storage device based at least in part on the determination that the attempt to obtain the first stripe metadata page was successful.

7. The apparatus of claim 1 wherein the at least one enclosure controller is further configured:

to add an additional entry to the metadata journal for an attempted write of a second stripe metadata page to the first metadata storage device based at least in part on the first metadata storage device being in the hiccup state.

8. The apparatus of claim 7 wherein in response to an attempted read of the second stripe metadata page from the first metadata storage device while the first metadata storage device is in the hiccup state, the at least one enclosure controller is further configured to read the second stripe metadata page from another of the metadata storage devices.

9. The apparatus of claim 7 wherein the at least one enclosure controller is further configured:

to determine that a threshold criterion is met by the first metadata storage device;

to set the first metadata storage device to a healthy state based at least in part on the determination that the threshold criterion is met by the first metadata storage device; and

to resynchronize the first metadata storage device based at least in part on the determination that the threshold criterion is met by the first metadata storage device.

10. The apparatus of claim 9 wherein resynchronizing the first metadata storage device comprises:

for each entry corresponding to the first metadata storage device in the metadata journal:

reading a valid copy of the corresponding first stripe metadata page from another of the metadata storage devices;

writing the valid copy of the corresponding first stripe metadata page to the first metadata storage device; and

invalidating the entry in the metadata journal.

11. A method comprising:

writing a first stripe metadata page to a plurality of metadata storage devices of a disk array enclosure, the first stripe metadata page corresponding to a stripe of data stored on a plurality of data storage devices of the disk array enclosure;

determining that the write of the first stripe metadata page failed for a first metadata storage device of the plurality of metadata storage devices;

adding an entry to a metadata journal stored in a cache of the disk array enclosure based at least in part on the determination that the write of the first stripe metadata page failed, the entry comprising an indication of the first metadata storage device and an indication of the stripe of data;

setting an indication in a data structure associated with the disk array enclosure that the first stripe metadata page for the stripe of data has not been written to the first metadata storage device; and

setting the first metadata storage device to a hiccup state based at least in part on the determination that the write of the first stripe metadata page failed for the first metadata storage device;

wherein the method is implemented by at least one enclosure controller of the disk array enclosure.

12. The method of claim 11 wherein the method further comprises:

determining that the write of the first stripe metadata page succeeded for a second metadata storage device of the plurality of metadata storage devices;

determining that the metadata journal comprises an additional entry corresponding to the second metadata storage device, the additional entry comprising an indication of the second metadata storage device and an indication of the stripe of data; and

invalidating the additional entry in the metadata journal based at least in part on the determination that the write of the first stripe metadata page succeeded for the second metadata storage device.

13. The method of claim 11 wherein the at least one enclosure controller comprises at least one processing device coupled to a memory, the memory comprising a dirty data structure associated with the first metadata storage device, the method further comprising adding an entry to the dirty data structure based at least in part on the determination that the write of the first stripe metadata page failed for the first metadata storage device, the entry added to the dirty data structure comprising a logical identifier corresponding to at least a portion of the stripe of data.

14. The method of claim 13 wherein the method further comprises:

receiving a read operation comprising the logical identifier;

determining that the dirty data structure associated with the first metadata storage device comprises the logical identifier; and

attempting to obtain the first stripe metadata page from another of the plurality of metadata storage devices based at least in part on the determination that the dirty data structure associated with the first metadata storage device comprises the logical identifier.

15. The method of claim 14 wherein:

the memory comprises a plurality of dirty data structures each associated with a corresponding metadata storage device of the plurality of metadata storage devices;

the method further comprises:

maintaining a copy of the first stripe metadata page on the plurality of data storage devices;

determining that each of the dirty data structures comprises the logical identifier; and

obtaining the first stripe metadata page from the plurality of data storage devices based at least in part on the determination that the each of the dirty data structures comprises the logical identifier.

16. The method of claim 14 wherein the method further comprises:

determining that the attempt to obtain the first stripe metadata page from the another of the plurality of metadata storage devices was successful; and

copying the first stripe metadata page to the first metadata storage device based at least in part on the determination that the attempt to obtain the first stripe metadata page was successful.

17. The method of claim 11 further comprising:

adding an additional entry to the metadata journal for an attempted write of a second stripe metadata page to the first metadata storage device based at least in part on the first metadata storage device being in the hiccup state.

18. The method of claim 17 wherein in response to an attempted read of the second stripe metadata page from the first metadata storage device while the first metadata storage device is in the hiccup state, the method further comprises reading the second stripe metadata page from another of the metadata storage devices.

19. The method of claim 17 wherein the method further comprises:

determining that a threshold criterion is met by the first metadata storage device;

setting the first metadata storage device to a healthy state based at least in part on the determination that the threshold criterion is met by the first metadata storage device; and

resynchronizing the first metadata storage device based at least in part on the determination that the threshold criterion is met by the first metadata storage device, the resynchronizing comprising:

for each entry corresponding to the first metadata storage device in the metadata journal:

reading a valid copy of the corresponding first stripe metadata page from another of the metadata storage devices;

writing the valid copy of the corresponding first stripe metadata page to the first metadata storage device; and

invalidating the entry in the metadata journal.

20. A computer program product comprising a non-transitory processor-readable storage medium having stored therein program code of one or more software programs, the program code when executed by at least one enclosure controller of a disk array enclosure of a storage system, causes the at least one enclosure controller:

to write a first stripe metadata page to a plurality of metadata storage devices of the disk array enclosure, the first stripe metadata page corresponding to a stripe of data stored on a plurality of data storage devices of the disk array enclosure;

to determine that the write of the first stripe metadata page failed for a first metadata storage device of the plurality of metadata storage devices;

to add an entry to a metadata journal stored in a cache of the disk array enclosure based at least in part on the determination that the write of the first stripe metadata page failed, the entry comprising an indication of the metadata storage device and an indication of the stripe of data;

to set an indication in a data structure associated with the disk array enclosure that the first stripe metadata page for the stripe of data has not been written to the first metadata storage device; and

to set the first metadata storage device to a hiccup state based at least in part on the determination that the write of the first stripe metadata page failed for the first metadata storage device.