作者：周志湖

主要内容

1. Spark Streaming与Kafka版的WordCount示例（一）
2. Spark Streaming与Kafka版的WordCount示例（二）

1. Spark Streaming与Kafka版本的WordCount示例 （一）

1. 启动kafka集群

root@sparkslave02:/hadoopLearning/kafka_2.10-0.8.2.1# bin/kafka-server-start.sh config/server.properties 
root@sparkslave01:/hadoopLearning/kafka_2.10-0.8.2.1# bin/kafka-server-start.sh config/server.properties 
root@sparkmaster:/hadoopLearning/kafka_2.10-0.8.2.1# bin/kafka-server-start.sh config/server.properties 

向kafka集群发送消息

root@sparkslave01:/hadoopLearning/kafka_2.10-0.8.2.1# bin/kafka-console-producer.sh --broker-list sparkslave01:9092 --sync --topic kafkatopictest

1. 编写如下程序

import org.apache.kafka.clients.producer.{ProducerConfig, KafkaProducer, ProducerRecord}
import org.apache.log4j.{Level, Logger}

import org.apache.spark.streaming._
import org.apache.spark.streaming.kafka._
import org.apache.spark.{Logging, SparkConf}

object KafkaWordCount {
  def main(args: Array[String]) {
    if (args.length < 4) {
      System.err.println("Usage: KafkaWordCount <zkQuorum> <group> <topics> <numThreads>")
      System.exit(1)
    }
    StreamingExamples.setStreamingLogLevels()

    val Array(zkQuorum, group, topics, numThreads) = args
    val sparkConf = new SparkConf().setAppName("KafkaWordCount").setMaster("local[4]")
    val ssc = new StreamingContext(sparkConf, Seconds(2))
    ssc.checkpoint("checkpoint")

    val topicMap = topics.split(",").map((_, numThreads.toInt)).toMap
    //创建ReceiverInputDStream
    val lines = KafkaUtils.createStream(ssc, zkQuorum, group, topicMap).map(_._2)
    val words = lines.flatMap(_.split(" "))
    val wordCounts = words.map(x => (x, 1L))
      .reduceByKeyAndWindow(_ + _, _ - _, Minutes(10), Seconds(2), 2)
    wordCounts.print()

    ssc.start()
    ssc.awaitTermination()
  }
}

配置运行参数：

具体如下：

sparkmaster:2181  test-consumer-group kafkatopictest 1

sparkmaster:2181，zookeeper监听地址
test-consumer-group， consumer-group的名称，必须和$KAFKA_HOME/config/consumer.properties中的group.id的配置内容一致
kafkatopictest，topic名称
1，线程数

运行KafkaWordCount 后，在producer中输入下列内容

root@sparkslave01:/hadoopLearning/kafka_2.10-0.8.2.1# bin/kafka-console-producer.sh --broker-list sparkslave01:9092 --sync --topic kafkatopictest
[2015-11-04 03:25:39,666] WARN Property topic is not valid (kafka.utils.VerifiableProperties)
Spark
Spark TEST
TEST Spark Streaming

得到结果如下：

2. Spark Streaming与Kafka版本的WordCount示例（二）

前面的例子中，producer是通过kafka的脚本生成的，本例中将给出通过编写程序生成的producer

// 随机生成1-100间的数字
object KafkaWordCountProducer {

  def main(args: Array[String]) {
    if (args.length < 4) {
      System.err.println("Usage: KafkaWordCountProducer <metadataBrokerList> <topic> " +
        "<messagesPerSec> <wordsPerMessage>")
      System.exit(1)
    }

    val Array(brokers, topic, messagesPerSec, wordsPerMessage) = args

    // Zookeeper连接属性配置
    val props = new HashMap[String, Object]()
    props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG, brokers)
    props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG,
      "org.apache.kafka.common.serialization.StringSerializer")
    props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG,
      "org.apache.kafka.common.serialization.StringSerializer")
    //创建KafkaProducer
    val producer = new KafkaProducer[String, String](props)

    // 向kafka集群发送消息
    while(true) {
      (1 to messagesPerSec.toInt).foreach { messageNum =>
        val str = (1 to wordsPerMessage.toInt).map(x => scala.util.Random.nextInt(10).toString)
          .mkString(" ")

        val message = new ProducerRecord[String, String](topic, null, str)
        producer.send(message)
      }

      Thread.sleep(1000)
    }
  }

}

KafkaWordCountProducer 运行参数设置如下：

sparkmaster:9092 kafkatopictest 5 8

sparkmaster:9092，broker-list
kafkatopictest，top名称
5表示每秒发多少条消息
8表示每条消息中有几个单词

先KafkaWordCountProducer，然后再运行KafkaWordCount ，得到的计算结果如下：