ZFS Faulted Pool, can't recover

none · May 11, 2018

Hi,

I have this pool and my box got a power outage. Then the pool never came back

I tried the clear and import commands:

Code:

root@xxx:~ # zpool import
   pool: wd750
     id: 9081684093828353742
  state: FAULTED
 status: The pool was last accessed by another system.
 action: The pool cannot be imported due to damaged devices or data.
        The pool may be active on another system, but can be imported using
        the '-f' flag.
   see: http://illumos.org/msg/ZFS-8000-EY
 config:

        wd750           FAULTED  corrupted data
          label/WD_750  ONLINE
root@xxx:~ # zpool import -F wd750
cannot import 'wd750': pool may be in use from other system, it was last accessed by yyy (hostid: 0x6e9a483c) on Thu Feb  8 22:48:25 2018
use '-f' to import anyway
root@xxx:~ # zpool import -f wd750
cannot import 'wd750': I/O error
        Destroy and re-create the pool from
        a backup source.
root@xxx:~ # zpool clear -F wd750
cannot open 'wd750': no such pool
root@xxx:~ # zpool import wd750
cannot import 'wd750': pool may be in use from other system, it was last accessed by yyy (hostid: 0x6e9a483c) on Thu Feb  8 22:48:25 2018
use '-f' to import anyway

Is there any kind of scrub or fsck I can do to it so I can at least read it? It is better to loose some files than all of them :/

thanks,

none

ShelLuser · May 12, 2018

Don't mess with commands such as clear if you don't know what you're doing. That is... if you value your data of course.

No, there is no such thing as fsck for ZFS.

Try: # zpool import -o readonly -fNR /mnt wd750. This might help you get access to the data.

Instead of messing with clear (I strongly suggest not to repeat this) I suggest checking -F instead. Careful with that though.
So.. what happens here?

Datapanic · May 12, 2018

zpool status

Sounds like your power fault knocked out 1 or more disks. Are they all up and spinning? Are they showing up in /dev? Need more info!

FAULTED is usually bad news in that there's "online" which is good, "degraded" which means data integrity is still there but you have 1 or more bad disks that need to be replaced, and finally, "FAULTED" which means the pool is not recoverable.

ralphbsz · May 12, 2018

How many disks are expected to be in the pool?

Why was this disk accessed by another computer?

none · May 12, 2018

ShelLuser said:
Don't mess with commands such as clear if you don't know what you're doing. That is... if you value your data of course.

No, there is no such thing as fsck for ZFS.

Try: # zpool import -o readonly -fNR /mnt wd750. This might help you get access to the data.

Instead of messing with clear (I strongly suggest not to repeat this) I suggest checking -F instead. Careful with that though.
So.. what happens here?

First of all, thanks. I did a whole disk dd backup beforehand. So that says why I have all that courage. I tried it, unfortunately the same outcome

Code:

[root@xxx /]# zpool import -o readonly -fNR /mnt/temp wd750
cannot import 'wd750': I/O error
        Destroy and re-create the pool from
        a backup source.

I never tried the -F here. Should I?

If I read the pool so I can take the files from there it will be perfect. Its all I want

thanks

Datapanic said:
zpool status

Sounds like your power fault knocked out 1 or more disks. Are they all up and spinning? Are they showing up in /dev? Need more info!

FAULTED is usually bad news in that there's "online" which is good, "degraded" which means data integrity is still there but you have 1 or more bad disks that need to be replaced, and finally, "FAULTED" which means the pool is not recoverable.

The disk is ok, all I think. It is on the /dev fine, and I got to dd if for a backup (I have the WD_750 label on it, that what I dd'ed after all).

Thanks

ralphbsz said:
How many disks are expected to be in the pool?

Why was this disk accessed by another computer?

No, it was not. This is a pool created and only used (at that point) on that box. A PI2 box. Now I is on another box running amd64 FreeBSD 11.1.

thanks to all,

none

PS: before reading here I was working on freeing a 1TB disk to use the dd'ed image on it. As I know little about ZFS internals, I was trying to create a zpool and try to force the bytes there somehow.

none · May 13, 2018

Update:

Tried the -F option as ShelLuser told, nothing.

Did the dd to a new disk and no good also. Guess I will have to forget about all the data now. Next time I will just use UFS2 disks for non-raid drives.

thanks,

none