Spark – Cyclic, high-performance data processing on top of Hadoop