Hey, I really hope someone can help me:

When the problem started a few days ago, first the shares disappeared, then everything froze. But a restart solved the whole thing again. Then the next time I saw the following in the logs:
"Jan 25 21:01:02 dry kernel: XFS (dm-0): Metadata corruption detected at xfs_inode_validate_extsize+0xb8/0x10b [xfs], inode 0x6022519f dinode
Jan 25 21:01:02 dry kernel: XFS (dm-0): Unmount and run xfs_repair
Jan 25 21:01:02 dry kernel: XFS (dm-0): First 128 bytes of corrupted metadata buffer:"
I ran XFS repair, then it ran again for a short time but then the problem came back, everything froze.
In the meantime the server hangs after a few minutes, so it is completely unusable.

I really hope someone can tell more from the logs.

 

dry-diagnostics-20220125-2102.zip

Check filesystem on disk1, make sure you run xfs_repair without -n or nothing will be done.

Ok, sorry for my late response. I run the xfs repair again, but the problem is still there. But I think there is also a new issue: the docker froze again, but this time the shares are still there and the GUI is still accesable. On the main Page it now shows this message: BTRFS operation is running

Should I let it run the BTRFS operation? And do you think that the error also comes from the filesystem metadata corruption?

 

PS: Diagnostic funktion also freezes, if I try to download it, so I added only the log

 

syslog.rtf.zip

Edited January 30, 20224 yr by desinox

4 hours ago, desinox said:

I run the xfs repair again, but the problem is still there.

Post the command you used and the full output of xfs_repair.

 

4 hours ago, desinox said:

so I added the only the log

Lots of crashing, you should run memtest.

I used the GUI option under "disk 1" in maintanance mode without the -n flag.

OK, I will try running memtest

You were totally right, Memtest spits out a lot of errors. I will try to find another stick.

After fixing the RAM run xfs_repair again, trying to fix it with bad RAM might makes things worse.