Major version upgrades on RDS are even more of a pain because you can't reliably plan the downtime you'll need for them to finish. You push a button and then you pray that it doesn't take much longer than your test runs with restored snapshots...

And then you still have to upgrade extensions and run analyze manually.

Aurora still can't do zero downtime upgrades though, which surprised me. You get closer with vanilla RDS PostgreSQL. And if you use RDS proxy or pgbouncer it's better.

At a previous company I had to tend to a large here of MySQL clusters and we did all sorts of zero-downtime stuff there. I've found MySQL replication far easier to work with.

It still doesn’t do zero downtime upgrades though, does it?