I\'m trying to get a structured streaming aggregation/groupBy working for kafka data. I\'m working on (Py)Spark 2.4.6. But I am getting empty output, no matter whichever mode.