Removing a stripe from a zfs pool¶
In my quest to optimize on power usage, One of my servers has a zfs pool consisting of 4x 8T disks, in striped mirrors configuration.
For this particular pool, IOPs and performance is not needed. As well, I am using barely 10% of the available space in this particular pool.
So- I decided to remove a pair of disks, to save a miniscule amount of energy.
Danger
These steps are intended for openzfs.
Also- I am not responsible for data loss. Don't perform actions on your pool, without tested, working backups.
(Snapshots / Raid) is not backups.
Step 1. Identify your pool¶
zpool status can be used to identify your pool, and its vdevs.
root@Tower:~# zpool status
pool: cache
state: ONLINE
scan: resilvered 1.59M in 00:00:00 with 0 errors on Wed Dec 13 19:09:21 2023
config:
NAME STATE READ WRITE CKSUM
cache ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
nvme2n1p1 ONLINE 0 0 0
nvme1n1p1 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
nvme0n1p1 ONLINE 0 0 0
nvme3n1p1 ONLINE 0 0 0
errors: No known data errors
pool: main
state: ONLINE
scan: scrub repaired 0B in 01:47:49 with 0 errors on Sun Dec 10 11:47:50 2023
config:
NAME STATE READ WRITE CKSUM
main ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
sdc1 ONLINE 0 0 0
sdf1 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
sdg1 ONLINE 0 0 0
sdi1 ONLINE 0 0 0
In my case, I want to remove the mirror-1 vdev from the main pool.
Step 2. Remove the vdev¶
Removing a top-level vdev is easy to do.
zpool remove main mirror-1
Where the syntax is- zpool remove pool-name vdev-name
Documentation for this command Can be found on the openzfs github
Running this command will not yield any output. However, you can view the current progress using zpool status
root@Tower:~# zpool status
pool: cache
state: ONLINE
scan: resilvered 1.59M in 00:00:00 with 0 errors on Wed Dec 13 19:09:21 2023
config:
NAME STATE READ WRITE CKSUM
cache ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
nvme2n1p1 ONLINE 0 0 0
nvme1n1p1 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
nvme0n1p1 ONLINE 0 0 0
nvme3n1p1 ONLINE 0 0 0
errors: No known data errors
pool: main
state: ONLINE
scan: scrub repaired 0B in 01:47:49 with 0 errors on Sun Dec 10 11:47:50 2023
remove: Evacuation of mirror in progress since Wed Dec 13 21:18:05 2023
6.06G copied out of 940G at 98.6M/s, 0.64% done, 2h41m to go
config:
NAME STATE READ WRITE CKSUM
main ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
sdc1 ONLINE 0 0 0
sdf1 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
sdg1 ONLINE 0 0 0
sdi1 ONLINE 0 0 0
errors: No known data errors
At this point, you just need to wait for the pool to finish moving data around.
It will yield the estimated time, and once completed, the vdev will be removed.
Step 3. Wait¶
At this point, you just need to wait for it to finish. Once finished, the VDEV will be automatically removed.
If- you use Unraid, you can also see the activity via its GUI. (At least, seeing reads from one stripe- and writes going to the other)
Using bash, you can use watch zpool status <poolname> which will display the current status every few seconds.
Every 2.0s: zpool status main Tower: Wed Dec 13 21:38:46 2023
pool: main
state: ONLINE
scan: scrub repaired 0B in 01:47:49 with 0 errors on Sun Dec 10 11:47:50 2023
remove: Evacuation of mirror in progress since Wed Dec 13 21:18:05 2023
150G copied out of 940G at 123M/s, 15.91% done, 1h49m to go
config:
NAME STATE READ WRITE CKSUM
main ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
sdc1 ONLINE 0 0 0
sdf1 ONLINE 0 0 0
mirror-1 ONLINE 0 0 0
sdg1 ONLINE 0 0 0
sdi1 ONLINE 0 0 0
errors: No known data errors