I have a QCOW2 image (Homeassistant VM), that I ran for several months without problems.
A few days ago, I reinstalled the VM host,so I copied the image to a backup drive and now wanted to start a VM from this image.
However, it always end up hanging at “booting from hard disk” and takes up 100% load on one core.
On the VM host, I imported the image like this:
# copied from HAOS wiki
sudo virt-install --name hass --description "Home Assistant OS" --os-variant=generic --ram=2048 --vcpus=2 --disk /var/vm/hass.qcow2,bus=sata --import --graphics none
To ensure that my host wasn’t broken, I tried the same image on another machine, that I know can run VMs (virtual machine manager, using the GUI), but same result. One core at 100% and no change at all.
I even let it run over night, but it was still at this point.
One machine runs NixOS, the other Debian 12.
What could cause this? There are no errors in journalctl or /var/log/qemu.
You must log in or register to comment.
Could you try addkng “-serial stdio” and see if any info comes out on the console?
You could also try debugging with gdb. https://qemu-project.gitlab.io/qemu/system/gdb.html
This may me way more technical than you wish to dive into, but it also might shed some light on where things are stuck.
I have also had VMs that hang on boot-up while running a core at 100%. I use Proxmox as my host, and was always able to enter the console for the VM and observe it going through the boot sequence. When it hung as you have described, it was because of some error that I could correct in the console. So, maybe check if your host allows you to interact with the VM while it’s booting.
I already tried that using virtual machine manager, I can see the “bios screen” and the blinking cursor after “booting from disk”, but that’s it.
So there is nothing happening, at least vm manager doesn’t even register any RAM usage, just one core at 100%.
Is your qcow2 image backed by another image?
Have you tried qemu-image check with -r for repair?