Robin Bloor wrote a nice post explaining how RainStor compresses data using tree structures and nor storing duplicate values.  Robin then goes on to explain how the RainStor architecture integrates well with Hadoop.  Definitely worth a read: RainStor and Hadoop.