Spark Apache Spark - Save Time with Less Join Operations Some of my surprising findings when optimizing Spark code