话不多说，直接上代码：

import org.apache.flink.api.common.typeinfo.TypeInformation
import org.apache.flink.api.java.typeutils.TypeExtractor
import org.apache.flink.streaming.api.TimeCharacteristic
import org.apache.flink.streaming.api.scala._
import org.apache.flink.streaming.connectors.kafka.{FlinkKafkaConsumer, KafkaDeserializationSchema}
import org.apache.kafka.clients.consumer.ConsumerRecord

import java.util.Properties

object Consumer {
  def main(args: Array[String]): Unit = {
    //创建flink环境
    val env = StreamExecutionEnvironment.getExecutionEnvironment
    //设置时间语义为，事件时间
    env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime)


    // kafka config
    val props = new Properties()
    props.setProperty("bootstrap.servers", "hosts") //hosts
    props.setProperty("group.id", "droupid") //消费者分组id
    props.setProperty("retries", "10") //重试次数
    props.setProperty("retries.backoff.ms", "100") //每次重试的间隔


    //kafka consumer
    val mykafka = new FlinkKafkaConsumer[User]("topicname", new MyKafkkaDes, props)


    //flink source
    val source = env.addSource(mykafka)

    source.print() //打印


    env.execute("jobname")

  }

}


case class User(id: Long, name: String, age: Int)

class MyKafkkaDes extends KafkaDeserializationSchema[User] {
  override def isEndOfStream(t: User): Boolean = false

  override def deserialize(consumerRecord: ConsumerRecord[Array[Byte], Array[Byte]]): User = {
    val str = new String(consumerRecord.value())
    val lists = str.split(",")

    val userid = lists(0).toLong
    val name = lists(1)
    val age = lists(2).toInt
    User(userid, name, age)
  }

  override def getProducedType: TypeInformation[User] = TypeExtractor.getForClass(classOf[User])
}