com.tresata.spark.kafka
Returns two Maps (PartitionId -> Offset) containing smallest and largests offsets for each partition*
Write contents of this RDD to Kafka messages by creating a Producer per partition.