【Spark】SparkStreaming从不同基本数据源读取数据

---

基本数据源

文件数据源

注意事项

1.SparkStreaming不支持监控嵌套目录
2.文件进入dataDirectory（受监控的文件夹）需要通过移动或者重命名实现
3.一旦文件移动进目录，则不能再修改，即使修改也不会读取修改后的数据

步骤

一、创建maven工程并导包

<properties>
    <scala.version>2.11.8scala.version>
    <spark.version>2.2.0spark.version>
properties>
<dependencies>
    <dependency>
        <groupId>org.scala-langgroupId>
        <artifactId>scala-libraryartifactId>
        <version>${scala.version}version>
    dependency>
    <dependency>
        <groupId>org.apache.sparkgroupId>
        <artifactId>spark-core_2.11artifactId>
        <version>${spark.version}version>
    dependency>
    <dependency>
        <groupId>org.apache.sparkgroupId>
        <artifactId>spark-sql_2.11artifactId>
        <version>${spark.version}version>
    dependency>
    
    <dependency>
        <groupId>org.apache.sparkgroupId>
        <artifactId>spark-streaming_2.11artifactId>
        <version>2.2.0version>
    dependency>
    <dependency>
        <groupId>org.apache.hadoopgroupId>
        <artifactId>hadoop-clientartifactId>
        <version>2.7.5version>
    dependency>

    <dependency>
        <groupId>org.apache.sparkgroupId>
        <artifactId>spark-hive_2.11artifactId>
        <version>2.2.0version>
    dependency>

    <dependency>
        <groupId>mysqlgroupId>
        <artifactId>mysql-connector-javaartifactId>
        <version>5.1.38version>
    dependency>

dependencies>
<build>
    <sourceDirectory>src/main/scalasourceDirectory>
    <testSourceDirectory>src/test/scalatestSourceDirectory>
    <plugins>
        <plugin>
            <groupId>org.apache.maven.pluginsgroupId>
            <artifactId>maven-compiler-pluginartifactId>
            <version>3.0version>
            <configuration>
                <source>1.8source>
                <target>1.8target>
                <encoding>UTF-8encoding>
                
            configuration>
        plugin>
        <plugin>
            <groupId>net.alchim31.mavengroupId>
            <artifactId>scala-maven-pluginartifactId>
            <version>3.2.0version>
            <executions>
                <execution>
                    <goals>
                        <goal>compilegoal>
                        <goal>testCompilegoal>
                    goals>
                    <configuration>
                        <args>
                            <arg>-dependencyfilearg>
                            <arg>${project.build.directory}/.scala_dependenciesarg>
                        args>
                    configuration>
                execution>
            executions>
        plugin>
        <plugin>
            <groupId>org.apache.maven.pluginsgroupId>
            <artifactId>maven-shade-pluginartifactId>
            <version>3.1.1version>
            <executions>
                <execution>
                    <phase>packagephase>
                    <goals>
                        <goal>shadegoal>
                    goals>
                    <configuration>
                        <filters>
                            <filter>
                                <artifact>*:*artifact>
                                <excludes>
                                    <exclude>META-INF/*.SFexclude>
                                    <exclude>META-INF/*.DSAexclude>
                                    <exclude>META-INF/*.RSAexclude>
                                excludes>
                            filter>
                        filters>
                        <transformers>
                            <transformer implementation="org.apache.maven.plugins.shade.resource.ManifestResourceTransformer">
                                <mainClass>mainClass>
                            transformer>
                        transformers>
                    configuration>
                execution>
            executions>
        plugin>
    plugins>
build>

二、在HDFS创建目录，并上传要做测试的数据

cd /export/servers/
vim wordcount.txt

hello world
abc test
hadoop hive

HDFS上创建目录

hdfs dfs -mkdir /stream_data
hdfs dfs -put wordcount.txt /stream_data

三、开发SparkStreaming代码

package cn.itcast.sparkstreaming.demo1

import org.apache.spark.streaming.dstream.DStream
import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.streaming.{Seconds, StreamingContext}

object getHdfsFiles {

  // 自定义updateFunc函数
  /**
   * updateFunc需要两个参数
   *
   * @param newValues    新输入数据计数累加的值
   * @param runningCount 历史数据计数累加完成的值
   * @return 返回值是Option
   *
   *         Option是scala中比较特殊的类，是some和none的父类,主要为了解决null值的问题
   */
  def updateFunc(newValues: Seq[Int], runningCount: Option[Int]): Option[Int] = {
    val finalResult: Int = newValues.sum + runningCount.getOrElse(0)
    Option(finalResult)
  }

  def main(args: Array[String]): Unit = {
    //获取SparkConf
    val sparkConf: SparkConf = new SparkConf().setAppName("getHdfsFiles_to_wordcount").setMaster("local[6]").set("spark.driver.host", "localhost")
    // 获取SparkContext
    val sparkContext = new SparkContext(sparkConf)
    // 设置日志级别
    sparkContext.setLogLevel("WARN")
    // 获取StreamingContext
    val streamingContext = new StreamingContext(sparkContext, Seconds(5))
    // 将历史结果都保存到一个路径下
    streamingContext.checkpoint("./stream.check")

    // 读取HDFS上的文件
    val fileStream: DStream[String] = streamingContext.textFileStream("hdfs://node01:8020/stream_data")
    // 对读取到的文件进行计数操作
    val flatMapStream: DStream[String] = fileStream.flatMap(x => x.split(" "))
    val wordAndOne: DStream[(String, Int)] = flatMapStream.map(x => (x, 1))
    // reduceByKey不会将历史消息的值进行累加，所以需要用到updateStateByKey，需要的参数是updateFunc，需要自定义
    val byKey: DStream[(String, Int)] = wordAndOne.updateStateByKey(updateFunc)

    //输出结果
    byKey.print()

    streamingContext.start()
    streamingContext.awaitTermination()
  }
}

四、运行代码后，往HDFS文件夹上传文件

五、控制台输出结果

-------------------------------------------
Time: 1586856345000 ms
-------------------------------------------

-------------------------------------------
Time: 1586856350000 ms
-------------------------------------------

-------------------------------------------
Time: 1586856355000 ms
-------------------------------------------
(abc,1)
(world,1)
(hadoop,1)
(hive,1)
(hello,1)
(test,1)

-------------------------------------------
Time: 1586856360000 ms
-------------------------------------------
(abc,1)
(world,1)
(hadoop,1)
(hive,1)
(hello,1)
(test,1)

-------------------------------------------
Time: 1586856365000 ms
-------------------------------------------
(abc,1)
(world,1)
(hadoop,1)
(hive,1)
(hello,1)
(test,1)

-------------------------------------------
Time: 1586856370000 ms
-------------------------------------------
(abc,2)
(world,2)
(hadoop,2)
(hive,2)
(hello,2)
(test,2)

-------------------------------------------
Time: 1586856375000 ms
-------------------------------------------
(abc,2)
(world,2)
(hadoop,2)
(hive,2)
(hello,2)
(test,2)

---

自定义数据源

步骤

一、使用nc工具给指定端口发送数据

nc -lk 9999

二、开发代码

import org.apache.spark.streaming.dstream.{DStream, ReceiverInputDStream}
import org.apache.spark.streaming.{Seconds, StreamingContext}
import org.apache.spark.{SparkConf, SparkContext}

object CustomReceiver {

  /**
   * 自定义updateFunc函数
   * @param newValues
   * @param runningCount
   * @return
   */
  def updateFunc(newValues:Seq[Int], runningCount:Option[Int]):Option[Int] = {
    val finalResult: Int = newValues.sum + runningCount.getOrElse(0)
    Option(finalResult)
  }

  def main(args: Array[String]): Unit = {
    // 获取SparkConf
    val sparkConf: SparkConf = new SparkConf().setAppName("CustomReceiver").setMaster("local[6]").set("spark.driver.host", "localhost")
    // 获取SparkContext
    val sparkContext = new SparkContext(sparkConf)
    sparkContext.setLogLevel("WARN")
    // 获取StreamingContext
    val streamingContext = new StreamingContext(sparkContext, Seconds(5))
    streamingContext.checkpoint("./stream_check")

    // 读取自定义数据源的数据
    val stream: ReceiverInputDStream[String] = streamingContext.receiverStream(new MyReceiver("node01", 9999))

    // 对数据进行切割、计数操作
    val mapStream: DStream[String] = stream.flatMap(x => x.split(" "))
    val wordAndOne: DStream[(String, Int)] = mapStream.map((_, 1))
    val byKey: DStream[(String, Int)] = wordAndOne.updateStateByKey(updateFunc)

    // 输出结果
    byKey.print()

    streamingContext.start()
    streamingContext.awaitTermination()
  }
}

import java.io.{BufferedReader, InputStream, InputStreamReader}
import java.net.Socket
import java.nio.charset.StandardCharsets

import org.apache.spark.storage.StorageLevel
import org.apache.spark.streaming.receiver.Receiver

class MyReceiver(host:String,port:Int) extends Receiver[String](StorageLevel.MEMORY_AND_DISK_2){
  /**
   * 自定义receive方法接收socket数据，并调用store方法将数据保存起来
   */
  private def receiverDatas(): Unit ={
    // 接收socket数据
    val socket = new Socket(host, port)
    // 获取socket数据输入流
    val stream: InputStream = socket.getInputStream
    //通过BufferedReader ，将输入流转换为字符串
    val reader = new BufferedReader(new InputStreamReader(stream,StandardCharsets.UTF_8))

    var line: String = null
    //判断读取到的数据不为空且receiver没有被停掉时
    while ((line = reader.readLine()) != null && !isStopped()){
      store(line)
    }

    stream.close()
    socket.close()
    reader.close()
  }


  /**
   * 重写onStart和onStop方法，主要是onStart，onStart方法会被反复调用
   */
  override def onStart(): Unit = {
    // 启动通过连接接收数据的线程
    new Thread(){
      //重写run方法
      override def run(): Unit = {
        // 定义一个receiverDatas接收socket数据
        receiverDatas()
      }
    }
  }

  // 停止结束的时候被调用
  override def onStop(): Unit = {

  }
}

---

RDD队列

步骤

一、开发代码

package cn.itcast.sparkstreaming.demo3

import org.apache.spark.rdd.RDD
import org.apache.spark.streaming.dstream.{DStream, InputDStream}
import org.apache.spark.{SparkConf, SparkContext}
import org.apache.spark.streaming.{Seconds, StreamingContext}

import scala.collection.mutable

object QueneReceiver {
  def main(args: Array[String]): Unit = {
    //获取SparkConf
    val sparkConf: SparkConf = new SparkConf().setMaster("local[6]").setAppName("queneReceiver").set("spark.driver.host", "localhost")
    //获取SparkContext
    val sparkContext = new SparkContext(sparkConf)
    sparkContext.setLogLevel("WARN")
    //获取StreamingContext
    val streamingContext = new StreamingContext(sparkContext, Seconds(5))

    val queue = new mutable.SynchronizedQueue[RDD[Int]]
    // 需要参数 queue: Queue[RDD[T]]
    val inputStream: InputDStream[Int] = streamingContext.queueStream(queue)
    // 对DStream进行操作
    val mapStream: DStream[Int] = inputStream.map(x => x * 2)

    mapStream.print()

    streamingContext.start()
    //定义一个RDD队列
    for (x <- 1 to 100){
      queue += streamingContext.sparkContext.makeRDD(1 to 10)
      Thread.sleep(3000)
    }
    streamingContext.awaitTermination()

  }
}