Year after year, I keep wondering (to put it lightly), who did name the restart function of virsh "destroy", and why.
It seems to me to be a complete -head thing to do; but it must have made sense to somebody, and I would like to understand the rationale.
The VM instance is forcefully shutdown, its memory freed. The VM configuration is not destroyed. The images/virtual disks/filesystems are not destroyed (except for possible corruption, as @DavidPostill pointed out).
So what is destroyed?
The KVM API itself uses restart/*"_RESTART", as one would expect, so the naming is not derived from the next layer.
Any insights to this?
Also, it is kept this way. Is there a conceptual idea behind that, or is this just kept for historical reasons?