The idea behind the ‘elastic’ cloud is that you can simply ’stretch’ your wasteband by adding more junk in your trunk, i.e. putting more servers online to give you the ability to scale.  The biggest problem here, however, is the simple fact that Postgres really operates on a ’shared-nothing’ architecture, so how is it that we gain anything by simply adding more resources?  That is, how does Postgres become elastic without changing the application or doing massive architecture work?

Right off the bat, I’ll attempt to eliminate confusion.  The following describes a feature specific to EnterpriseDB’s proprietary product, Postgres Plus Advanced Server, and, specifically, InfiniteCache.

(more…)