How to spread receivers over worker hosts evenly in Spark streaming In Spark Streaming, you can spawn multiple receivers to increase parallelism, e.g., such that each receiver reads from one of the partitions in Kafka. Then you combine the resulting streams and process them by batches. …