question of my own: when y'all are performance testing postgres, what size of DB instance are you using? we're seeing some bad behavior from an rds db.t4g.micro in a system with ~100k relation tuples and are trying to figure out whether our sizing is wayyy off