Hadoop’s Overhead – Our Experiments with Big Data