People in research groups, what is your setup for shared server/resources?

So my university finally decided to shell out some money to meet their expectations of our research scaling up to, *gulp*, "big data". I am not sure as to what we are going to get (they do not want to give too much money either, since it is mainly for us lowly PhD students) but their idea is that we should use it for sharing datasets and running experiments. We are a mainly text mining group.

I do not know much about hardware stuff but I am pretty sure we are going to need a system for people to submit jobs to the server to be run sequentially, so that it doesn't die from overuse. Also, we kind of talked about writing a set of libraries to stream datasets from the server to individual machines for people wanting to run their experiments locally, but not wanting to download all the data on their machine.

Does anybody know of prepackaged software we could use for those two things? What is your setup? We do 99% of our programming in Java if that matters.