Here’s a summary of what we tried during the troubleshooting session for the issue I was having with Boxes:
I first tried to open the problematic virtual machine (VM) in Boxes, but it closed immediately without any error message.
We checked the system logs to see if there were any relevant errors, but found nothing.
I tried running the sudo boxes -l command to list the VMs, but it didn’t work.
We discovered that Boxes uses libvirt to manage VMs, and that we could use the virsh command to interact with libvirt.
I checked the status of the libvirtd service using sudo systemctl status libvirtd, and it was running fine.
I tried running sudo virsh list --all, but it didn’t show any VMs.
We checked the system logs again using sudo journalctl -u libvirtd.service, but didn’t find anything useful.
I tried running sudo virsh --connect qemu:///system dumpxml fedora37-wor-3-3, but got an error saying that the domain didn’t exist.
We checked the folder where the VM images are stored using ls ~/.local/share/gnome-boxes/images/, and found that the images were there.
I ran sudo qemu-img info on the image files to check their format and size.
I tried starting the problematic VM using sudo gnome-boxes start fedora37-wor-3-3, but it failed with the same issue.
We tried running sudo ausearch -c ‘qemu-kvm’ --raw | audit2allow -M my-qemukvm and sudo semodule -X 300 -i my-qemukvm.pp to see if SELinux was causing the issue, but they didn’t find any issues.
So far, we have been unable to determine the cause of the issue with Boxes not being able to open the VM.