In the previous article I gave the background to a project we did for a client, exploring the benefits of Spark-based ETL processing running on Amazon's »