I am doing an experiment here with wordcount using pull methods with Spark Streaming (2.2.0) and Flume (1.6.0). Things are working fine on my local VM, so I move everything onli