Great points in this article - though using something like cascalog makes hadoop suck a lot less - e.g composable more complex queries closer in power to sql. W

Great points in this article - though using something like cascalog makes hadoop suck a lot less - e.g composable more complex queries closer in power to sql. Wouldn't be quite as crazy to use on smaller datasets be it just for fun or to prove you are ready if/when your dataset grows large enough.