Make sure that in case of malfunction we have multiple fail-safe ways of gaining access to the PHX production VMs to resolve issues.
can help with this task?
I don't think this epic is relevant since we have OpenVPN in PHX so SPICE consoles work from any location as long as the OpenVPN machine is running.
If I'm misunderstanding we should define what exactly is missing currently and see if it can be implemented securely.
We only have one access option, if SPICE is not working for some reason - we have no backup.
We need to ensure the other options are available as well, at least VNC, and having the console which would probably be way faster then both would be useful as well.