大数据-Spark的任务执行

Spark的任务执行

Spark-submit

（1）修改conf/slaves配置文件

hadoop1

（2）启动spark的伪分布式集群

./sbin/start-all.sh

（3）spark-submit提交任务（以蒙特卡洛求圆周率为例）

spark-submit --master spark://hadoop1:7077 --class org.apache.spark.examples.SparkPi /usr/local/spark/spark-2.1.0-bin-hadoop2.7/examples/jars/spark-examples_2.11-2.1.0.jar 100

（4）spark-submit运行结果

Spark-shell

本地模式

（1）修改conf/slaves配置文件

hadoop1

（2）启动spark的伪分布式集群

./sbin/start-all.sh

（3）启动spark-shell

spark-shell

（4）提交任务

sc.textFile("spark_workCount.txt").flatMap(_.split(" ")).map((_,1)).reduceByKey(_+_).collect

集群模式

（1）修改conf/slaves配置文件

hadoop1

（2）启动spark的伪分布式集群

./sbin/start-all.sh

（3）启动spark-shell

spark-shell --master spark://hadoop1:7077

（4）hdfs创建/spark/tmp文件夹

hdfs dfs -mkdir -p /spark/tmp

（5）hdfs上传spark_workCount.txt文件

hdfs dfs -put spark_workCount.txt /spark/tmp

（6）提交任务

sc.textFile("hdfs://hadoop1:9000/spark/tmp/spark_workCount.txt").flatMap(_.split(" ")).map((_,1)).reduceByKey(_+_).saveAsTextFile("hdfs://hadoop1:9000/spark/output")

Spark的WordCount

Scala本地模式

（1）将jars文件夹下的jar包放到IDEA项目的resource下

注意：保持scala版本跟jar包的版本一致

（2）启动spark-shell

./sbin/start-all.sh
spark-shell

（3）运行本地模式的WorkCount

package spark

import org.apache.spark.{SparkConf, SparkContext}

object WordCount {
  def main(args: Array[String]): Unit = {
    //创建一个spark的配置文件
    val conf = new SparkConf().setAppName("Scala WorkCount").setMaster("local")
    //实例化SparkContext对象
    val sc = new SparkContext(conf)

    //本地模式
    val result = sc.textFile("hdfs://192.168.138.130:9000/spark/tmp/spark_workCount.txt")
      .flatMap(_.split(" "))
      .map((_,1))
      .reduceByKey(_+_)

    //输出结果
    result.foreach(println)
  }
}

Scala集群模式

（1）编写scala代码

import org.apache.spark.{SparkConf, SparkContext}

object WorkCount {
  def main(args: Array[String]): Unit = {
    //创建一个spark的配置文件
    val conf = new SparkConf().setAppName("Scala WorkCount")
    //实例化SparkContext对象
    val sc = new SparkContext(conf)

    //集群模式
    val result = sc.textFile(args(0))
      .flatMap(_.split(" "))
      .map((_,1))
      .reduceByKey(_+_)
      .saveAsTextFile(args(1));

    //关闭
    sc.stop();
  }
}

（2）将代码打成jar包放到linux上

（3）运行spark

./sbin/start-all.sh

（4）提交任务

./sbin/start-all.sh
spark-submit --master spark://hadoop1:7077 --class spark.WordCount /root/Spark-1.0-SNAPSHOT.jar hdfs://192.168.138.130:9000/spark/tmp/spark_workCount.txt hdfs://192.168.138.130:9000/spark/wordcount

Java本地模式

（1）将jars文件夹下的jar包放到IDEA项目的resource下

注意：保持scala版本跟jar包的版本一致

（2）启动spark-shell

./sbin/start-all.sh
spark-shell

（3）运行本地模式的WorkCount

package com.spark.util;

import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.JavaSparkContext;
import org.apache.spark.api.java.function.FlatMapFunction;
import org.apache.spark.api.java.function.Function2;
import org.apache.spark.api.java.function.PairFunction;
import scala.Tuple2;

import java.util.Arrays;
import java.util.Iterator;
import java.util.List;

/**
 * Spark WordCount
 * 
 * @author Jabin
 * @version 1.00 2019/
 */
public class WordCount {
    public static void main(String[] args) {
        //创建Spark配置
        SparkConf conf = new SparkConf().setAppName("Spark.WordCount").setMaster("local");
        //加载Spark配置
        JavaSparkContext sc = new JavaSparkContext(conf);
        //本地模式
        JavaRDD<String> textFile = sc.textFile("hdfs://192.168.138.130:9000/spark/tmp/spark_workCount.txt");

        JavaRDD<String> flatMap = textFile.flatMap(new FlatMapFunction<String, String>() {
            public Iterator<String> call(String s) {
                return Arrays.asList(s.split(" ")).iterator();
            }
        });

        JavaPairRDD<String, Integer> map = flatMap.mapToPair(new PairFunction<String, String, Integer>() {
            public Tuple2<String, Integer> call(String s) {
                return new Tuple2<String, Integer>(s, 1);
            }
        });

        JavaPairRDD<String, Integer> reduce = map.reduceByKey(new Function2<Integer, Integer, Integer>() {
            public Integer call(Integer a, Integer b) {
                return a + b;
            }
        });

        List<Tuple2<String, Integer>> list = reduce.collect();

        for (Tuple2<String, Integer> tuple: list){
            System.out.println(tuple._1+" : "+tuple._2);
        }
    }
}

Spark的任务调度架构图