Comment by acdha

What's the ops experience for a replicated setup like these days? i.e. assuming you want basic fault-tolerance at non-exotic size / activity levels, how much of a job is someone acquiring if, say, there are reasons they can't just use AWS RDS?