I know this is the opposite around. Most people here are asking about importing pools. I am however finding that exporting pools is impossible.
I had two pools Storj-1 and Storj-2 that I am no longer using. I wish to export them in a tidy manner so I can shutdown the expansion tray.
From the GUI I get pool is busy
From the CLI I get pool is busy
Wether I destroy or export
Nothing is (should be) using those pools. I could just pull the disks out / switch off the disk tray - but I am now curious. I have had this issue before but always solved it by just rebooting, after which the pools can generally be removed. In this case however I am:
- In the middle of a big replication - so do not want to reboot
- Curious
[Update] I just opened the CLI and cd to /mnt and neither Storj-1 or Storj-2 are present implying that as far as ZFS is concerned the pools are no longer present. Its a shame that the GUI thinks they are still in existence.
Do you see any open files?
lsof | grep "/mnt/<poolname>"
There are no files - see my latest update above
lsof just generates what looks like a load of errors about no pwd entry for UID 911 (and others)
Are the pools actually imported and active? Maybe the GUI is lying.
zpool status
Hadn’t though of that
pool: Storj-1
state: ONLINE
scan: scrub repaired 0B in 1 days 01:25:52 with 0 errors on Mon Jun 2 01:25:59 2025
config:
NAME STATE READ WRITE CKSUM
Storj-1 ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
72833db1-92f2-4826-b38f-c032c336f8ec ONLINE 0 0 0
30f5ee67-c985-4d51-96af-449c8f4ef553 ONLINE 0 0 0
errors: No known data errors
pool: Storj-2
state: ONLINE
scan: scrub repaired 0B in 1 days 04:48:11 with 0 errors on Tue Jun 3 04:48:21 2025
config:
NAME STATE READ WRITE CKSUM
Storj-2 ONLINE 0 0 0
mirror-0 ONLINE 0 0 0
19506586-5838-41b6-b2c8-037568c92316 ONLINE 0 0 0
e7cccc1e-2a27-4730-b34d-0d5939fccc6f ONLINE 0 0 0
So yes - the pools are present, but not
I suppose you could export them with the command-line. You might get the same “pool is busy” message.
Tried that - and yes got pool is busy
Looks like you’re not using hot spares either.
You can either attempt with -f
or try to unmount the datasets, one by one, if you’re really curious.
You’re going to wipe these disks anyways?
Tried -f
There are no datasets left (other than the root). There is no data left that I have any interest in
No shares actively using them. No hot spares. No open files. I’m assuming you’re not cd’d into one of their paths.
One of the world’s greatest mysteries.
They don’t have a path any longer - thats caught me out before
Pool names are no longer in /mnt
A reboot will I believe fix things so the pools can be removed from the GUI and from zpool. I just can’t reboot atm