Cache drive suddenly unmountable/unsupported.

 

Tried stopping array, unassigning, reassigning etc.

 

Cache drive apparently is associated with my Docker so none of those show up now.

tower-diagnostics-20241215-1627.zip

Try typing:

 

btrfs rescue zero-log /dev/sde1

 

Then restart the array and post new diags.

New logs here

tower-diagnostics-20241216-1012.zip

Btrfs is detecting data corruption on the pool, recommend running memtest, then scrub the pool

8 hours ago, JorgeB said:

Btrfs is detecting data corruption on the pool, recommend running memtest, then scrub the pool

 

So doing that I can now see the drive on the main tab without "unmountable: unsupported or no file system" text.

I can also access the files that i presume were still on there/was on cache and mover hadnt ran. This was via windows though.

The scrub results FYI:
Scrub started: Mon Dec 16 13:50:20 2024

Status: finished Duration: 0:03:30

Total to scrub: 90.15GiB

Rate: 524.78MiB/s

Error summary: csum=98 Corrected: 0 Uncorrectable: 98 Unverified: 0

 

 

Now all my dockers arent visible and im presented with "Docker Service failed to start."

This is my main issue, I dont mind losing some genuine "cache" drive download files but my docker setup i want to fix.

 

Thanks for your help so far.

 

 

Not clear if you have already run memtest? If there are RAM issues, try to continue may cause more damage, but to see the docker issue post new diags.

14 hours ago, JorgeB said:

Not clear if you have already run memtest? If there are RAM issues, try to continue may cause more damage, but to see the docker issue post new diags.

Apologies.

 

I was doing all this remotely and couldnt test MEMTEST.

I have returned now and can confirm 1 of the 2 sticks in my system returns errors.

I am continuing with a single stick for now.

 

New diags attached.

tower-diagnostics-20241217-0935.zip

Start by running a correcting scrub for the pool, and post the results.

 

9 minutes ago, JorgeB said:

Start by running a correcting scrub for the pool, and post the results.

 

UUID: 540d6691-560b-44cd-99b5-11eda8d4250c

Scrub started: Tue Dec 17 10:05:47 2024

Status: finished Duration: 0:02:56

Total to scrub: 90.15GiB Rate: 524.49MiB/s

Error summary: csum=99 Corrected: 0 Uncorrectable: 99 Unverified: 0

Look in the syslog for the list of corrupt files and delete them or restore from a backup, then run another scrub to confirm 0 errors.

On 12/17/2024 at 11:20 AM, JorgeB said:

Look in the syslog for the list of corrupt files and delete them or restore from a backup, then run another scrub to confirm 0 errors.

Thanks again,

 

I can go to the sys log and looking at each of the log types am I right in saying its the "warnings" im interested in?

 

As that the only place i can see lines like

"Dec 17 10:07:51 Tower kernel: BTRFS warning (device sde1): checksum error at logical 104288641024 on dev /dev/sde1, physical 89264644096, root 5, inode 4074617, offset 373710848, length 4096, links 1 (path: data/media/movies/******/********.mp4)"

 

I haven no problem with just losing these data/movie files.

 

But i dont see anything listed related to Docker/my containers which is all that i am interested in.

 

Sorry, quite a noob at this.

 

 

Edited December 19, 20241 yr by ItzMar

37 minutes ago, ItzMar said:

log types am I right in saying its the "warnings" im interested in?

Yes, they should be logged during the scrub, after deleting/restoring, run another scrub to confirm 0 errors.

Im confused, i dont see anything about my docker containers in the logs only data files.

 

Then if i try to delete those data files i cannot, i see nobody is the owner, but it wont allow me to change the owner either.

 

That means those files are still in the pool, the mover will fail to move if they are corrupt, but you should be able to delete them manually, try using the CLI.