Zpool import issues after upgrade from 9.2 to 9.3 (fresh install) due to dead USB boot stick

[moppidoo]

Hi,
As the title says, I'm having a bit of an issue after the installation when importing the ZFS volumes. after scouring through the forum threads for hours trying to find answers, I have came to some conclusion but also technical hurdle which I cannot figure out how to get over.

I have this one particularly overfilled volume prior to the replacement/upgrade of FreeNAS and now have trouble importing via GUI, the GUI sees the volume and always end up with errors and asked to check the pool status.

So I opened up the shell and do a zpool status and got the below output (truncated):

Code:

pool: WiredSTO_02                                                                                                                
state: ONLINE                                                                                                                     
  scan: scrub repaired 0 in 10h50m with 0 errors on Sat Jan 10 16:20:32 2015                                                       
config:                                                                                                                            
                                                                                                                                   
        NAME                                            STATE     READ WRITE CKSUM                                                 
        WiredSTO_02                                     ONLINE       0     0     0                                                 
          raidz2-0                                      ONLINE       0     0     0                                                 
            gptid/f24bc20e-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f2bfda19-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f33d08f8-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f3beaa89-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f46a91fd-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f5067074-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f59ddfa2-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
            gptid/f6094fa9-441e-11e2-9642-bc5ff448f101  ONLINE       0     0     0                                                 
                                                                                                                                   
errors: No known data errors

so I exported the pool in CLI and attempted a manual mount with

Code:

zpool import -R /mnt WiredSTO_02

and got the following result:

Code:

cannot mount '/mnt/WiredSTO_02/MMWPublic_02': failed to create mountpoint                                                          
cannot mount '/mnt/WiredSTO_02/P_Kollectionz': failed to create mountpoint

using:

Code:

zfs list

returns (truncated):

Code:

NAME USED AVAIL REFER MOUNTPOINT
WiredSTO_02                                                      10.1T      0   427K  /mnt/WiredSTO_02                             
WiredSTO_02/MMWPublic_02                                         4.04T      0  4.04T  /mnt/WiredSTO_02/MMWPublic_02                
WiredSTO_02/P_Kollectionz                                        6.08T      0  6.08T  /mnt/WiredSTO_02/P_Kollectionz

I read people have success deleting/nulling non-important files in the pool before they are able to Auto Import the volume via GUI again in several threads, however, my question is, all my files are in the 2 dataset mountpoints that failed to mount and I don't quite know how to access them so that I can delete the files.

The other potential option is to add extra disk/vdev to the pool which might solve the problem, however what do I need for that to happen, this is a 8x2TB RAIDZ2 array, and I don't quite know how to achieve it with the GUI even assuming if I have the hardware resources to do it as the volume won't show up in the Volume Manager.

Any help would be appreciated!

moppidoo

 

[cyberjock]

So it looks like your dataset metadata is likely corrupted beyond the ability to be mounted. At this point I have no advice except to restore from backup. There's a reason why you get that 80% full warning, then the red 95% full. It's actually dangerous because you can lose the data forever