How to use kafka.group.id in spark 3.0 to avoid always start the offset for latest position?

来源：https://stackoverflow.com/questions/64003405/how-to-use-kafka-group-id-in-spark-3-0-to-avoid-always-start-the-offset-for-late