Found this interesting set of videos about distributed computing using the hadoop project by Apache and a distributed filesystem. Its the sort of technology google use and is the technology that facebook use (plus many others). It allows you to run jobs over many hundred computers and process terabyes of data. There is also a database-esk layer that sits on top of this which allows you to do things like joins on really large datasets just like in sql.
http://vimeo.com/3579013