You would think that after the previous big us-east-1 outages (to be fair there have been like 3 of them in the past decade, but still, that's plenty), companies would have started to move to other AWS regions and/or to spread workloads between them.
It’s not that simple. The bit AWS doesn’t talk much about publicly (but will privately if you really push them) is that there’s core dependencies behind the scenes on us-east-1 for running AWS itself. When us-east-1 goes down the blast radius has often impacted things running in other regions.
It impacts AWS internally too. For example rather ironically it looks like the outage took out AWS’s support systems so folks couldn’t contact support to get help.
Unfortunately it’s not as simple as just deploying in multiple regions with some failover load balancing.
our eu-central-1 services had zero disruption during this incident, the only impact was that if we _had_ had any issues we couldn't log in to the AWS console to fix them
So moving stuff out of us-east-1 absolutely does help
Sure it helps. Folks just saying there’s lots of examples where you still get hit by the blast radius of a us-east-1 issue even if you’re using other regions.
