the reason I'm asking I've been playing with the read/write endpoints on a few puppet resources and foreman to see if I can optimise read actions and write actions against RDS, rather than scaling the whole instance, deploying and scaling cluster read or writer resources