Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Friends,

had one hard drive fail in a 4x1.5 TB ZFS setup and am now stuck. status output is below. Since it's a raidz1 vdev, I am still hoping for recovering it. My issue is that since I cannot import the pool, I cannot 'zpool detach' or 'zpool destroy' the now non-existent device - only get 'pool is unavailable'. That happens even when I import with the secrect -V option (only way I don't get an error message in the first place).

I have searched every forum, but to no avail. Anybody got any ideas remaining? I am hoping that by removing the faulted disk from the configuration I will go to degraded status, but simply cannot do that.

Thanks a million, Alex

#zpool status
pool: tank state: FAULTED status: One or more devices could not be used because the label is missing or invalid. There are insufficient replicas for the pool to continue functioning. action: Destroy and re-create the pool from a backup source. see: http://www.sun.com/msg/ZFS-8000-5E scrub: none requested config:

  NAME                      STATE     READ WRITE CKSUM
   tank                      FAULTED      0     0     1  corrupted data
     raidz1-0                DEGRADED     0     0     6
       c1t1d0s1              ONLINE       0     0     0
       15212365201588890516  FAULTED      0     0     0  was /dev/dsk/c1t1d0s1
       c1t3d0s1              ONLINE       0     0     0
       c1t2d0s1              ONLINE       0     0     0

#zpool detach tank 15212365201588890516
cannot open 'tank': pool is unavailable

#zpool list

NAME SIZE ALLOC FREE CAP DEDUP HEALTH ALTROOT

tank - - - - - FAULTED -

share|improve this question
    
Are you sure it's just one disk that has issues? The whole pool shouldn't go into the faulted state when just one disk drops. –  Shane Madden Jan 12 '13 at 20:56
    
Did you replace the bad disk with another? What OS are you running? –  ewwhite Jan 12 '13 at 21:12
    
I cannot replace the disk because the zfs replace command does not recognize the pool as existing - even after the -V import. Exactly the same as detach. As for the other disks, I can dd from them easily, is there a way to test whether their zfs metadata is fine? I have tried this on the eon storage appliance (which is based on Solaris snv_130) and under zevo on Mac OS. Thanks! –  Alex Niemeyer Jan 13 '13 at 14:20

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.