Clear, simple and agnostic; looking forward to a similar article with dual servers showing the failover, and then the next logical step of virtualizing the whole thing
