Strange crash/freeze of whole host caused by a VM, behavior continues on new host

xE1H

New Member
Jul 8, 2023
1
0
1
Hey, I'm experiencing really strange behavior. Basically, once in a while my host would freeze (gets question marks on the dashboard), sometimes displaying a "disk IO error" or something around the lines of "disk in read only mode" on all of the VM's stdin (just blasted out into the shell, as text input). I thought it's a disk issue, so I took the host down, but I found nothing that was inherently wrong. SMART readings were also fine, so I just moved the most important VM (my web VM, runs 6 gunicorn servers and NGINX Proxy Manager on a Docker container) to a new host. Fast forward to today, the new host froze in the same fashion.

Disks are not full, SMART readings are looking fine. They are practically brand new. I am 99.99% sure that it's not the connection or anything similar.

TL;DR: Host freezes displaying a disk error, behavior continues on new host after transfering VM.

Has anybody experienced something like this? How can I investigate further? Thanks in advance.

According to Grafana, the host went down 17:05:40.

Here's the /var/log/syslog of the host (lolek) and the vm (kvintas), nothing interesting. I restarted the machine @ 20:25.

1688839884814.png
1688839932472.png
 
Last edited: