> For HVM VMs, the running qemu/bhyve process sends an ACPI signal to the
> guest kernel telling it to shut down. In case the guest kernel ignores
> this or for some reason does not receive this request we mark the VM
> with a transition property indicating that we tried to shut it down.
> This transition marker also includes a timeout (default 180 seconds).
> If we hit the timeout, the VM is forcibly halted.
OpenBSD defaults to 30 seconds to stop all VMs before going postal and
terminating processes unless things have changed since 7.5.
/etc/rc runs "/etc/rc.d/vmd stop" before it runs rc.shutdown so I can't
automate my own VM shutdowns in rc.shutdown. (unless I edit /etc/rc)
Only approved solution is to up the timeout with "rcctl set vmd timeout"
to cover the time needed for all VMs to shutdown.
I also moved my VMs disk images to a new partitions mounted with "sync"
so as to not corrupt (as bad) the VM's disk during a power loss event.
My server doesn't like cheap UPSes and my wallet doesn't like expensive.
Eric
No comments:
Post a Comment