This week covered three topics: "how do you endure failure, how do you recover from disaster, and how do you move workloads." On the surface they look separate, but the three converge on one question — "what is the worst this system must survive, and how much will you spend to survive it." Is it a single-AZ failure or a whole-region blackout? Can you not lose even one second of data, or is an hour tolerable? Is what blocks the move downtime or bandwidth? Reading these constraints precisely and mapping them to the right-cost tool is the essence of SAA Domain 2 (Resilience)