Three ZFS mirrors disapeared after the system hanged and restarted.

Hi.
Today, while I was watching a movie streamed from emby, all froze and OpenMediaVault restarted.
When I could log in again all my mirrors was gone.
Only the OMV system disc is left.
I tried to import all pools, but got an exception:
exception 'OMVModuleZFSException' with message 'cannot mount '/media/puh': directory is not empty' in /usr/share/omvzfs/Utils.php:540

There are a lot of shared folders and other things pointing to these three mirrors, and I guess that the disks are fine, they have "just" dropped out of OMV.

But how do I go about to get it all back as before?

What logs can I produce to get the help I need to restore my system?

I had started to move files to empty one of the mirrors, so I could switch to xfs instead, as my system usually used over 90% of the 8Gb ram I have (and that is max).

Even if the data there can be restored from other sources, that will be a cumbersome task that I wish I can avoid...

Thanks. First I restarted OMV once, but still all was gone.
But now, after 30 to 45 minutes the shares have reappeared.
It might have helped to turn off all services as well, actually.
- As I got them back after I turned them all off (or I just noticed them after that).

So I had a scare, but it seems alright again now.

I take that back. Most of my volumes are back. One is still faulty.
That one is split into two devices, one is mounted the other is not.
And it is the device I got the error on earlier, "media/puh".

OK, actually read some more on the error:

exception 'OMVModuleZFSException' with message 'cannot import 'puh': a pool with that name already exists
use the form 'zpool import <pool | id> <newpool>' to give it a new name' in /usr/share/omvzfs/Utils.php:540

So it gives more info to me, but how do I browse to /usr/share/omvzfs/Utils.php? looks like a webpage URL, but i get a 404 if i try and browse to it.
I suspect that "zpool import" means the "import pool" button, that makes sense, but to enter "puh nalle" or "puh | nalle" only gives me "The pool already exist" or "No such pool available". Not sure what <pool | id> suggests (other than the name). I have the name, but not any ID for it, not that I know of any way.

I tried it in putty as well:

> zpool import puh nalle
cannot import 'puh': no such pool available

Thanks. Makes sense.

I tried 'zpool import puh' but got back "cannot import 'puh': a pool with that name already exists"

I guess that the issue lies with the existance of the item 'puh' in the File system page.
I cannot mount or unmount that one, I suspect I need to go in to aa file and remove it...

:~# zpool status
  pool: ior
 state: ONLINE
  scan: none requested
config:


        NAME        STATE     READ WRITE CKSUM
        ior         ONLINE       0     0     0
          mirror-0  ONLINE       0     0     0
            sda     ONLINE       0     0     0
            sde     ONLINE       0     0     0


errors: No known data errors


  pool: puh
 state: ONLINE
  scan: scrub repaired 0 in 1h18m with 0 errors on Sat Apr 23 07:31:55 2016
config:


        NAME        STATE     READ WRITE CKSUM
        puh         ONLINE       0     0     0
          mirror-0  ONLINE       0     0     0
            sdc     ONLINE       0     0     0
            sdd     ONLINE       0     0     0


errors: No known data errors


  pool: tiggr
 state: ONLINE
  scan: none requested
config:


        NAME        STATE     READ WRITE CKSUM
        tiggr       ONLINE       0     0     0
          mirror-0  ONLINE       0     0     0
            sdb     ONLINE       0     0     0
            sdf     ONLINE       0     0     0


errors: No known data errors

I created these pools on this system, but I have have done it a couple of times before I (almost) got it the way I want (the subfolders on the zfs is overkill).

Thanks!
That solved it (or 'zpool export -f puh' did, - I got an error with -f before the command)

My disc is back online.

@raulfg3 I have now reinstalled OMV to a new HD instead of the flashdrive.
The zfs drives work fine after import.

--Except that I need to import them after every reboot. That does not feel good...

I tried to export them all, delete the folders and then import them again, with the commands above. No luck.

Is there any system files to check to verify or to force add them?

Or a log that could help with troubleshooting this?

Thanks for your help.
It was not.

drwxr-xr-x 5 root root 5 Apr 23 15:02 ior
drwxrwxr-x 6 root users 6 May 13 15:11 puh
drwxr-xr-x 4 root root 4 Apr 22 16:27 tiggr

All these three are zpools. None of them are mounted automatically.
But as you see, all three have errors, two have 755 and the only one with 775 have group Users.

So I should be OK if I change the right for these three to root:root 775?

@raulfg3

That did not help. I still have to import them all three after restart.

Zitat von raulfg3

If finally you find how to solve, please share here, I want to know.

I found this thread on something similar - but for Ubuntu. https://github.com/zfsonlinux/zfs/issues/703

I tried to look at the files referenced but didn't find anything that seemed to fit with OMV.
Is there something there that can be applied on my situation?

As a workaround, can I use something like this in crontab?
@reboot sleep 10 ; zpool import -a -f

Or should I avoid that? It would be best to find a proper solution...

Edit: I was about to move the server to it's original space after the reinstallation.
After the restart I actually got one of the pools online. Degraded, but online. The others came up after import.
I check the reason for why it was degraded, and then shut it down. I fixed the loosened sata cable and rebooted.
This time the other two pools came online... The last one came online after import (all OK now).
Do not really know what you can do with this, I just add it for completion...