that would be pretty cool but to have that, you need a high-network-latency solution, i.e., pretty much cold back-up. For some time I thought it's pretty last century option but having been experimenting for some time now, it's the option with lowest impact on system performance. More importantly, it's reasonably resilient.
I've read your comment now about 4 times and all I have come up with is "huh?"
Literally thousands if not millions of organisations operate multi-DC infrastructure across the planet.
Is it harder than setting up a single box in one DC? Yes.
Is it harder than setting up a mini-cluster of boxes in one DC? Yes.
Is it rocket science? No.