org.apache.spark.sql.kafka010.KafkaDontFailOnDataLossSuite.failOnDataLoss=false should not return duplicated records: microbatch v1