I spent several years at AWS and now run my own (tiny)technical organization and we are all-in on AWS. I can’t even begin to justify multi-region failover, let alone multi-provider. Now that I’m in the startup ecosystem, I saw many others complain about this outage. I just want to shout at them ‘Do you have any idea how many $/engineers/engineering hours you save each and every day by using AWS?!?’
Outages happen. It sucks. The alternative is increasing your engineering cost AND slow down your entire development pipeline by some ridiculous factor just to stay up during those single-digit hours per year that AWS sees catastrophic failure. NO THANK YOU!
I know how seriously AWS takes these kinds of errors such that every year these sorts of issues occur less and less frequently. They are doing a fantastic job and I intend to keep tying myself to them.