I've gotten the impression that clustered Postgres still isn't very straightforward to run. Do you mind elaborating on your ideal setup and point to some resources?
It's not straightforward at all but it's better than it was five years ago ... you can use something more "meta" like SymmetricDS (https://www.symmetricds.org/). I haven't used it personally but a dirt simple way to get an HA, scalable PostgreSQL instance would be to use Amazon's Aurora DB.