

It does something similar, in that you can replicate servers and have one act as a failover for another, but I think you’d need a lot of extra code to sync the primary game host with the failover hosts, and more to make sure all the clients detect a failed host and all gracefully switch to the same failover host, and probably a hundred other things I’m not thinking of. If I were going to build something like this I probably would look to kubernetes as a foundation but there would be a lot of customization on top of that.
They ran the numbers, it’s actually cheaper to replace any equipment that gets damaged or killed.