Data just vanished ...
Posted: Mon Oct 19, 2020 10:56 pm
No idea what happened ... one minute, all is running fine, the next ... my data is gone. Or rather, the data appears to still be there, but something somehow corrupted the root of the volume ... ?
There *used* to be a few directories at the top, more important being: /Volumes/Plex/Seasons
Now, this is probably my fault, and am going to have to work the next while recovering the data ( possible, just very painful ), but after a reboot, I had 6 CKSUM errors on the middle volume:
Without thinking of checking /Volume/Plex, I ran zpool clear Plex since everything looked to be fine ... the reason why I power cycled the machine was due to that middle drive appearing to power off ( its in a RAID enclosure, so not sure how that could have happened, but neither here nor there ) ... when I powered back up, all three drives were found and running ...
I didn't create the files listed above, so it *looks* like the non .* directories were removed, but not the .* ones ...
From df, it looks like the underlying files still exist, just not the file name to match the inode at the top of it all ... ?
Is there anything that I can do / check to get that 6.3TB of data accessible again? Or is it just lost?
I haven't upgrade anything recently, but am running Catalina 10.15.7 with
And ZetaWatch r43
Nothing critical on those drives, but if I can somehow avoid rebuilding what was on there that would be optimal ...
Have checked what log files in /var/log I can find, nothing is jumping out at me ...
Thanks for any help ... am not touching anything on that file system until I hear back ;(
- Code: Select all
/dev/disk10s1 7.0Ti 6.3Ti 753Gi 90% 10759 197507213 0% /Volumes/Plex
- Code: Select all
# ls -l /Volumes/Plex
total 591
-rw-r--r--@ 1 scrappy staff 8196 Oct 19 23:30 .DS_Store
drwx------ 4 root wheel 4 Apr 2 2020 .Spotlight-V100
d-wx--x--t 5 root wheel 5 Jul 4 14:42 .Trashes
-rw-r--r-- 1 root wheel 239735 Apr 2 2020 .VolumeIcon.icns
drwx------ 29 root wheel 29 Oct 19 23:37 .fseventsd
There *used* to be a few directories at the top, more important being: /Volumes/Plex/Seasons
Now, this is probably my fault, and am going to have to work the next while recovering the data ( possible, just very painful ), but after a reboot, I had 6 CKSUM errors on the middle volume:
- Code: Select all
# zpool status -v Plex
pool: Plex
state: ONLINE
scan: scrub repaired 0 in 0 days 00:00:03 with 0 errors on Mon Oct 19 23:34:19 2020
config:
NAME STATE READ WRITE CKSUM
Plex ONLINE 0 0 0
raidz1-0 ONLINE 0 0 0
media-42DDE7B7-C87E-064B-A6B8-02107230D7AD ONLINE 0 0 0
media-167372F3-4258-FF4D-B26E-EA6E0A47C565 ONLINE 0 0 0
media-C29C64DE-EF26-FB47-BD29-8C789CF57000 ONLINE 0 0 0
errors: No known data errors
Without thinking of checking /Volume/Plex, I ran zpool clear Plex since everything looked to be fine ... the reason why I power cycled the machine was due to that middle drive appearing to power off ( its in a RAID enclosure, so not sure how that could have happened, but neither here nor there ) ... when I powered back up, all three drives were found and running ...
I didn't create the files listed above, so it *looks* like the non .* directories were removed, but not the .* ones ...
From df, it looks like the underlying files still exist, just not the file name to match the inode at the top of it all ... ?
Is there anything that I can do / check to get that 6.3TB of data accessible again? Or is it just lost?
I haven't upgrade anything recently, but am running Catalina 10.15.7 with
- Code: Select all
# zfs version
zfs-1.9.4-0
zfs-kmod-1.9.4-0
And ZetaWatch r43
Nothing critical on those drives, but if I can somehow avoid rebuilding what was on there that would be optimal ...
Have checked what log files in /var/log I can find, nothing is jumping out at me ...
Thanks for any help ... am not touching anything on that file system until I hear back ;(