Spark-Streaming核心编程

以下是今天所学的知识点与代码测试：

Spark-Streaming

DStream实操

案例一：WordCount案例

需求：使用 netcat 工具向 9999 端口不断的发送数据，通过 SparkStreaming 读取端口数据并统计不同单词出现的次数

实验步骤：

1. 添加依赖

<dependency>
    <groupId>org.apache.spark</groupId>
    <artifactId>spark-streaming_2.12</artifactId>
    <version>3.0.0</version>
</dependency>

1. 编写代码

val sparkConf = new SparkConf().setMaster("local[*]").setAppName("streaming")
val ssc = new StreamingContext(sparkConf,Seconds(3))

val lineStreams = ssc.socketTextStream("node01",9999)
val wordStreams = lineStreams.flatMap(_.split(" "))
val wordAndOneStreams = wordStreams.map((_,1))
val wordAndCountStreams = wordAndOneStreams.reduceByKey(_+_)
wordAndCountStreams.print()

ssc.start()
ssc.awaitTermination()

1. 启动netcat发送数据

nc -lk 9999

 Spark-Streaming核心编程（一）

DStream 创建

创建DStream的三种方式：RDD队列、自定义数据源、kafka数据源

RDD队列

可以通过使用 ssc.queueStream(queueOfRDDs)来创建 DStream，每一个推送到这个队列中的 RDD，都会作为一个DStream 处理。

案例：

需求：循环创建几个 RDD，将 RDD 放入队列。通过 SparkStream 创建 Dstream，计算 WordCount

代码：

val sparkConf = new SparkConf().setMaster("local[*]").setAppName("RDDStream")
val ssc = new StreamingContext(sparkConf, Seconds(4))

val rddQueue = new mutable.Queue[RDD[Int]]()
val inputStream = ssc.queueStream(rddQueue,oneAtATime = false)
val mappedStream = inputStream.map((_,1))
val reducedStream = mappedStream.reduceByKey(_ + _)
reducedStream.print()

ssc.start()
for (i <- 1 to 5) {
  rddQueue += ssc.sparkContext.makeRDD(1 to 300, 10)
  Thread.sleep(2000)
}
ssc.awaitTermination()

结果展示：

自定义数据源

自定义数据源需要继承 Receiver，并实现 onStart、onStop 方法来自定义数据源采集。

案例：自定义数据源，实现监控某个端口号，获取该端口号内容。

1. 自定义数据源

class CustomerReceiver(host:String,port:Int) extends Receiver[String](StorageLevel.MEMORY_ONLY) {
  override def onStart(): Unit = {
    new Thread("Socket Receiver"){
      override def run(): Unit ={
        receive()
      }
    }.start()
  }

  def receive(): Unit ={
    var socket:Socket = new Socket(host,port)
    var input :String = null
    var reader = new BufferedReader(new InputStreamReader(socket.getInputStream,StandardCharsets.UTF_8))

    input = reader.readLine()
    while(!isStopped() && input != null){
      store(input)
      input = reader.readLine()
    }
    reader.close()
    socket.close()

    restart("restart")
  }

  override def onStop(): Unit = {}
}

1. 使用自定义的数据源采集数据

val sparkConf = new SparkConf().setMaster("local[*]").setAppName("stream")
val ssc = new StreamingContext(sparkConf,Seconds(5))

val lineStream = ssc.receiverStream(new CustomerReceiver("node01",9999))

val wordStream = lineStream.flatMap(_.split(" "))
val wordAndOneStream = wordStream.map((_,1))
val wordAndCountStream = wordAndOneStream.reduceByKey(_+_)
wordAndCountStream.print()

ssc.start()
ssc.awaitTermination()