Problem/Justification
If multiple disks are replaced, sequential resilvering may be used, but oftentimes it is desired to perform a parallel resilver.
The only way to do this currently is to issue a zpool resilver <pool> command from the shell.
This was discussed in this thread
The below zpool status shows the result when a replacement is awaiting a previous resilver operation
# zpool status tank
pool: tank
state: ONLINE
status: One or more devices is currently being resilvered. The pool will
continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
scan: resilver in progress since Thu Oct 24 14:10:40 2024
1.84T / 20.9T scanned at 13.8G/s, 0B / 19.4T issued
0B resilvered, 0.00% done, no estimated completion time
config:
NAME STATE READ WRITE CKSUM
tank ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
1b2a34e0-e57c-4a5b-9033-ca0fe03f51a3 ONLINE 0 0 0
26609e36-6148-48b2-9c12-309a385d17a6 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
a167e02d-4382-4b04-9010-31036b3a29b5 ONLINE 0 0 0
37d7938e-d4dd-4e94-8005-7a4358c4ee17 ONLINE 0 0 0
mirror-2 ONLINE 0 0 0
85648b65-c328-4e25-b243-2cc2b9a6952e ONLINE 0 0 0
46b01e54-2d59-4eb4-97a3-3629d6fcd48b ONLINE 0 0 0
mirror-3 ONLINE 0 0 0
replacing-0 ONLINE 0 0 0
3048fff0-c7b0-4c84-98b4-b9b0d19f3999 ONLINE 0 0 0
b9bef6ea-b826-4e5c-80cf-d18c8ff930ba ONLINE 0 0 0
aa20557c-b462-4129-a45c-a6f859a4bbdd ONLINE 0 0 0
mirror-4 ONLINE 0 0 0
replacing-0 ONLINE 0 0 0
28ccfe09-ff3d-4095-9e55-7c3686dc18ef ONLINE 0 0 0
a2a37f56-b979-4658-bf02-6ec89f90d1fb ONLINE 0 0 0 (awaiting resilver)
8038480a-73f3-43d6-b4af-2601fcb7cf86 ONLINE 0 0 0
mirror-5 ONLINE 0 0 0
8b6c50d7-8227-4175-a41b-25bf3f8f49e7 ONLINE 0 0 0
sdk2 ONLINE 0 0 0
mirror-6 ONLINE 0 0 0
cb39e5a9-1c39-4a74-96d7-662644a4df9f ONLINE 0 0 0
sdu2 ONLINE 0 0 0
mirror-7 ONLINE 0 0 0
sdg2 ONLINE 0 0 0
1c11a1f1-678a-4b7c-a2ab-0feaec78acb2 ONLINE 0 0 0
mirror-8 ONLINE 0 0 0
1f18dc41-5935-4059-823b-bf0874742deb ONLINE 0 0 0
sdi2 ONLINE 0 0 0
mirror-15 ONLINE 0 0 0
2de3cf07-af92-4ba6-ba1d-b5da5f5316de ONLINE 0 0 0
bf819af9-1504-4d59-bc4a-c65606c0f0fd ONLINE 0 0 0
logs
nvme0n1p1 ONLINE 0 0 0
cache
sdw1 ONLINE 0 0 0
sdx1 ONLINE 0 0 0
errors: No known data errors
The key is that the second disk replacement is awaiting the first replacement to complete before it will start, even though the second disk is a separate vdev.
ZFS does this so as to not have to restart the initial resilver, which may have been running for hours or days… but of course, its only been running for a minute or so…
Additionally, TrueNAS gets confused when trying to figure out the ETA in this situation…
Issuing a manual resilver command will trigger the parallel resilver
# zpool resilver tank
# zpool status tank
pool: tank
state: ONLINE
status: One or more devices is currently being resilvered. The pool will
continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
scan: resilver in progress since Thu Oct 24 14:18:48 2024
3.47T / 20.9T scanned at 27.6G/s, 0B / 18.3T issued
0B resilvered, 0.00% done, no estimated completion time
config:
NAME STATE READ WRITE CKSUM
tank ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
1b2a34e0-e57c-4a5b-9033-ca0fe03f51a3 ONLINE 0 0 0
26609e36-6148-48b2-9c12-309a385d17a6 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
a167e02d-4382-4b04-9010-31036b3a29b5 ONLINE 0 0 0
37d7938e-d4dd-4e94-8005-7a4358c4ee17 ONLINE 0 0 0
mirror-2 ONLINE 0 0 0
85648b65-c328-4e25-b243-2cc2b9a6952e ONLINE 0 0 0
46b01e54-2d59-4eb4-97a3-3629d6fcd48b ONLINE 0 0 0
mirror-3 ONLINE 0 0 0
replacing-0 ONLINE 0 0 0
3048fff0-c7b0-4c84-98b4-b9b0d19f3999 ONLINE 0 0 0
b9bef6ea-b826-4e5c-80cf-d18c8ff930ba ONLINE 0 0 0
aa20557c-b462-4129-a45c-a6f859a4bbdd ONLINE 0 0 0
mirror-4 ONLINE 0 0 0
replacing-0 ONLINE 0 0 0
28ccfe09-ff3d-4095-9e55-7c3686dc18ef ONLINE 0 0 0
a2a37f56-b979-4658-bf02-6ec89f90d1fb ONLINE 0 0 0
8038480a-73f3-43d6-b4af-2601fcb7cf86 ONLINE 0 0 0
mirror-5 ONLINE 0 0 0
8b6c50d7-8227-4175-a41b-25bf3f8f49e7 ONLINE 0 0 0
sdk2 ONLINE 0 0 0
mirror-6 ONLINE 0 0 0
cb39e5a9-1c39-4a74-96d7-662644a4df9f ONLINE 0 0 0
sdu2 ONLINE 0 0 0
mirror-7 ONLINE 0 0 0
sdg2 ONLINE 0 0 0
1c11a1f1-678a-4b7c-a2ab-0feaec78acb2 ONLINE 0 0 0
mirror-8 ONLINE 0 0 0
1f18dc41-5935-4059-823b-bf0874742deb ONLINE 0 0 0
sdi2 ONLINE 0 0 0
mirror-15 ONLINE 0 0 0
2de3cf07-af92-4ba6-ba1d-b5da5f5316de ONLINE 0 0 0
bf819af9-1504-4d59-bc4a-c65606c0f0fd ONLINE 0 0 0
logs
nvme0n1p1 ONLINE 0 0 0
cache
sdw1 ONLINE 0 0 0
sdx1 ONLINE 0 0 0
errors: No known data errors
Notice the second replace is no longer awaiting a resilver
Ideally, the “Scrub” button which is disabled during a “resilver” operation, could perhaps be overloaded to “force resilver” or, perhaps if there is an “awaiting resilver” some button to start all resilvers could appear/apply?
I find that wanting to have multiple resilvers running is a fairly common operation, especially when upgrading disks…