Stav dette: Optimizing hadoop for MapReduce :