Docker crashes, doesn't start and need to reboot system

Diags appear to be after a reboot, when it happens again post new diags before rebooting.

 

Also see that btrfs is detecting data corruption in the pool:

 

Apr 14 16:58:16 Tower kernel: BTRFS info (device sdb1): bdev /dev/sdb1 errs: wr 0, rd 0, flush 0, corrupt 526, gen 0

 

Run a scrub and delete/restore from a backup the corrupt files, also a good idea to run memtest

I had run MemTest without any issues:

 

I'll post diagnostics when crashes again.

1 hour ago, JorgeB said:

Run a scrub and delete/restore from a backup the corrupt files,

Also make sure you do the above, and reset the btrfs stats to see if new corruptions are detected.

On 4/17/2024 at 4:55 PM, JorgeB said:

Also make sure you do the above, and reset the btrfs stats to see if new corruptions are detected.

New diagnostics uploaded, after docker crashes, beforr restart.

tower-diagnostics-20240421-1412.zip

Btrfs is detecting new data corruption, since memtest is only definitive if it finds errors, and since you have multiple sticks, fix the corrupt files and reset the stats and and try with just one stick of RAM, if the same try with a different one, that will basically rule out bad RAM.

On 4/22/2024 at 10:28 AM, JorgeB said:

Btrfs is detecting new data corruption, since memtest is only definitive if it finds errors, and since you have multiple sticks, fix the corrupt files and reset the stats and and try with just one stick of RAM, if the same try with a different one, that will basically rule out bad RAM.

 

new problems happen

tower-diagnostics-20240714-1151.zip

Server ran out of RAM, it appears to be caused by the container which is starting all theses processes:

 

Jul 14 01:17:03 Tower kernel: [  17461]  1000 17461  8475627     3963   663552        0           200 chromium
Jul 14 01:17:03 Tower kernel: [  17465]  1000 17465  8473539     3593   651264        0           200 chromium
Jul 14 01:17:03 Tower kernel: [  17480]  1000 17480 296529358    17344  1032192        0           300 chromium
Jul 14 01:17:03 Tower kernel: [  17482]  1000 17482 296529074     7857   868352        0           300 chromium
Jul 14 01:17:03 Tower kernel: [  17484]  1000 17484 296527041     5723   782336        0           300 chromium
Jul 14 01:17:03 Tower kernel: [  17540]  1000 17540  8473129     5707   688128        0           200 chromium
Jul 14 01:17:03 Tower kernel: [  17562]  1000 17562 296526822     4187   729088        0           300 chromium

 

Well, i think that is Krusader, thats the only one chromium based.

 

Thanks once more @JorgeB