Spark with Scala reading/writing files to HDFS with automatic additions of new Spark workers using Docker "scale"