I don't understand this paradigm I guess? For a small single server or development environment... I hate having to load 100's of thousands of records just to analyze it in a graph... am I missing the big picture here?
UPDATE (3/21/2012 10:38a):
My current setup:
How do you instantiate your database, embedded or server? Are you running ImpermanentGraphDatabase, because that's the in-memory test database. If you use the normal EmbeddedGraphDatabase your graph is persisted trasactionally along the way when you insert your data.
Please give a little more information.