Hi All,

New to unRAID and have been spending the last few weeks getting everything setup and my data moved over with great success.  It's been going really well and I'm really liking the product!  That being said, as of this morning I'm getting just unending errors like this: Jul 20 15:15:10 unRAID kernel: BTRFS critical (device sdi1): corrupt node, bad key order: block=164267491328, root=1, slot=44

I'm trying to stop the array, but it won't stop so I guess I'm just not really sure what the next steps are for me to get this resolved.  I'm attaching the diagnostics.  Any help or pointers would be greatly appreciated!

 

Thanks,

Matt

 

unraid-diagnostics-20180720-1515.zip

Edited July 20, 20187 yr by mbezzo
typos

alright, so it seems there's unrecoverable errors on the cache drive.  how can I delete my cache and recreate it?

You can run a scrub, if there are uncorrectable errors then reformat the pool.

thanks - scrub can't fix a few, can you point me to directions on reformatting the pool? I can't seem to find that piece!

 

Thanks!
 

appdata share is so messed up.  If I mount it and look around there's only a few of my docker container's directories.  BUT, if I use command line to cd to a docker that isn't visible... it works.  For some, not all.  Any tricks on how to recover?  I'm assuming it's gonna be nuke and start over...  I so don't understand what happened here but man, a painful lesson!

 

root@unRAID:/mnt/user/appdata# ls -al
total 36
drwxrwxrwx  1 nobody users  228 Jul 20 22:39 ./
drwxrwxrwx  1 nobody users  100 Jul 20 17:59 ../
-rw-rw-rw-+ 1 nobody users 6148 Jul 21 12:33 .DS_Store
-r--r--r--+ 1 nobody users    0 Jul  1 13:04 .com.bombich.ccc.casetest
-r--r--r--+ 1 nobody users    0 Jun 30 14:23 .com.bombich.ccchelper.casetest
drwxrwxrwx  1 nobody users   35 Jul  1 09:58 DDClient/
drwxrwxrwx  1 nobody users   18 Jun 24 12:39 DiskSpeed/
drwxrwxrwx  1 nobody users   54 Jul 21 09:17 binhex-radarr/
drwxrwxrwx  1 nobody users   80 Jul 20 14:22 binhex-sonarr/
drwxrwxrwx  1 nobody users    6 Jun 30 13:32 data/
drwxrwxr-x  1 nobody users   78 Jul 20 22:39 krusader/
drwxrwxrwx  1 nobody users  197 Jul 20 22:09 letsencrypt/
drwxrwxrwx  1 nobody users    6 Jul 20 22:07 letsencryptold/
drwxrwxrwx  1 nobody users   33 Jun 30 09:59 nzbget/
drwxrwxrwx  1 nobody users   88 Jul 20 22:52 ombi/
drwxrwxrwx  1 nobody users   20 Jul  2 17:35 organizr/
drwxrwxrwx  1 nobody users   24 Jul 21 12:31 tautulli/
root@unRAID:/mnt/user/appdata# cd PlexMediaServer
root@unRAID:/mnt/user/appdata/PlexMediaServer# ls -al
total 0
drwxrwxrwx 1 nobody users  14 Jul 16 21:00 ./
drwxrwxrwx 1 nobody users 228 Jul 20 22:39 ../
drwxr-xr-x 1 root   root   38 Jul 16 21:00 Library/

How is this even possible!?

 

 

If it is nuke and start over - what's the best way to do that? Simply delete the appdata share and recreate it?  

 

 

Appreciate the help!

You can try btrfs restore if you need to save some data.

 

When all data is backed up just wipe all the disks with blkdiscard, stop the array and run:

 

blkdiscard /dev/sdX

 

Replace X with each device, then start the array and reformat the pool.

thank you!