Cool. I'm keen to try it. Wondering how well it works with multi-terabyte data sets.