Change trove instance to be arbitrary container

Modify the instance object to make it a container for trove information and move instance information into a subresource called "nodes". A trove instance then becomes a container that allows us to better describe things like clusters or complex datastores/data processing tools like hadoop or cassandra.