[gfo]

Agreed. If they go as far as randomly disabling AWS nodes to test their own resilience I'd be shocked if they didn't have some sort of contingency for if AWS went completely belly-up. Who knows, they could even have Azure nodes on standby in case AWS in its entirety went down?

[sitkack]

Everyone should have a cloud provider total-outtage strategy. You should atleast control the ability to move DNS if everything in said cloud provider is out.