Forgot RAM was overclocked used to be a windows server. So I'll turn off XMP profile and then wipe and rebuild cache from scratch (you said backup and recreate, but if I just restored a backup wouldn't I reapply corruptions?) and I should hopefully not have issues again?
So I just checked the disk in maintenance mode with the GUI tool of btrfs check --readonly. Then ran it with --repair and it seems to have repaired successfully. Let me know if this was a bad idea?
So I'm completely unsure of what's going on. I think my NVME drive might be failing? The steps I've already taken are:
Delete docker.img and rebuild from there
Full memtest for 24 hours to be sure it wasn't memory error
Balanced NVME drive.
The system starts functioning again after reboot but a few hours later it fails again.
Most recent logs attached. I have previous ones also.
tower-diagnostics-20220122-2010.zip