This should always be a business calculation but I do think you should note that there’s at least an order of magnitude difficulty increase between multiple AZs and regions, especially if you’re using services like RDS where it’s designed in, so I’d consider that a solid bridge step.
I’m trying to put some numbers into that, I’ve been running a relatively well trafficked website in multiple AZs since 2011. We had ~20 minutes of downtime when they had a network routing issue for us-east-1 and a few hours of degraded service when S3 had a region-wide outage. I haven’t added up the number of single AZ outages during that period but based on the RSS feeds I think it’s a good bit more relative to the very modest additional cost.
Good point, the multi-AZ RDS feature is a nice way to get most of the resilience upsides without any additional app complexity. You do double your database cost, but that might be worth it.
That’s assuming that you a) religiously test with the smaller size and b) are comfortable that scaling up will work when lots of other people are shifting workloads, too. I usually work on projects where we haven’t wanted to deal with that but that’s a judgement call.
I’m trying to put some numbers into that, I’ve been running a relatively well trafficked website in multiple AZs since 2011. We had ~20 minutes of downtime when they had a network routing issue for us-east-1 and a few hours of degraded service when S3 had a region-wide outage. I haven’t added up the number of single AZ outages during that period but based on the RSS feeds I think it’s a good bit more relative to the very modest additional cost.