Describir: Using Spark in the Hadoop ecosystem :