将应用程序提交到spark环境并将结果输出到hdfs_spark dataframe输出到hdfs源码-CSDN博客

本文链接：https://blog.csdn.net/dream_broken/article/details/78520315

我们在“hadoop学习1--hadoop2.7.3集群环境搭建” “spark学习1--centOS7.2下基于hadoop2.7.3的spark2.0集群环境搭建” 中已经将hadoop、spark的集群环境都搭建起来了，jdk用的是1.7版本的。

1.启动hadoop集群

centOS7服务器3台

master 192.168.174.132

node1 192.168.174.133

node2 192.168.174.134

启动这3台虚拟机，在这3台虚拟机上执行jps，发现都还没启动hadoop

在master上启动hadoop

看启动日志，发现node1,node2也跟着启动了，在master,node1，node2虚拟机上执行jps会发现相关的进程启动了

[root@master sbin]# jps
2354 NameNode
2539 SecondaryNameNode
2947 Jps
2690 ResourceManager

浏览器访问http://192.168.174.132:8088/

2.上传单词文件到hadoop

master上执行

[root@master /]# hadoop fs -ls /
[root@master /]#

还没有任何目录，创建个目录，用来存放单词文件的原始文件

[root@master /]# hadoop fs -mkdir /wc_input
[root@master /]# hadoop fs -ls /
Found 1 items
drwxr-xr-x   - root supergroup          0 2017-11-13 15:17 /wc_input
[root@master /]#

使用工具把windows上的words.txt文件上传到master跟目录

[root@master /]# hadoop fs -put /words.txt /wc_input
[root@master /]# hadoop fs -ls /wc_input
Found 1 items
-rw-r--r--   2 root supergroup       2190 2017-11-13 15:20 /wc_input/words.txt
[root@master /]#

3.启动spark集群

在master上执行

[root@master /]# $SPARK_HOME/sbin/start-all.sh
starting org.apache.spark.deploy.master.Master, logging to /soft/spark/spark-2.0.0-bin-hadoop2.7/logs/spark-root-org.apache.spark.deploy.master.Master-1-master.out
node2: starting org.apache.spark.deploy.worker.Worker, logging to /soft/spark/spark-2.0.0-bin-hadoop2.7/logs/spark-root-org.apache.spark.deploy.worker.Worker-1-node2.out
node1: starting org.apache.spark.deploy.worker.Worker, logging to /soft/spark/spark-2.0.0-bin-hadoop2.7/logs/spark-root-org.apache.spark.deploy.worker.Worker-1-node1.out

浏览器访问http://192.168.174.132:8080/

3.代码打包

将代码稍微修改下，能读取外面传进去的参数；由于spark环境搭建的时候，jdk用的是1.7版本，所以不能用lambda表达式。

package com.fei;

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

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;

/**
 * 单词统计，并按降序排序，输出前10个单词及个数
 * @author Jfei
 *
 */
public class WordCount {

	public static void main(String[] args) {
		System.out.println(args[0]);
		System.out.println(args[1]);
		System.out.println(args[2]);
		//1.本地模式，创建spark配置及上下文
	//	SparkConf conf = new SparkConf().setAppName("wordCount").setMaster("spark://192.168.174.132:7077");
		SparkConf conf = new SparkConf().setAppName("wordCount").setMaster(args[0]);
		
		JavaSparkContext sc = new JavaSparkContext(conf);
		
		//2.读取本地文件,并创建RDD
	//	JavaRDD<String> linesRDD = sc.textFile("e:\\words.txt");
	//	JavaRDD<String> linesRDD = sc.textFile("hdfs://192.168.174.132:9000/wc_input");
		JavaRDD<String> linesRDD = sc.textFile(args[1]);
		//3.每个单词由空格隔开,将每行的linesRDD拆分为每个单词的RDD
	//	JavaRDD<String> wordsRDD = linesRDD.flatMap(s  -> Arrays.asList(s.split("\\s")));
		//相当于 ==>
		JavaRDD<String> wordsRDD = linesRDD.flatMap(new FlatMapFunction<String, String>(){
			private static final long serialVersionUID = 1L;
			@SuppressWarnings("unchecked")
			@Override
			public Iterator<String> call(String line) throws Exception {
				return Arrays.asList(line.split(" ")).listIterator();
			}
		});
		//4.将每个单词转为key-value的RDD，并给每个单词计数为1
		//JavaPairRDD<String,Integer> wordsPairRDD = wordsRDD.mapToPair(s -> new Tuple2<String,Integer>(s, 1));
		//相当于 ==>
		JavaPairRDD<String,Integer> wordsPairRDD = wordsRDD.mapToPair(new PairFunction<String, String, Integer>() {
			private static final long serialVersionUID = 1L;
			@Override
			public Tuple2<String, Integer> call(String word) throws Exception {
				return new Tuple2<String,Integer>(word,1);
			}
		});
		
		//5.计算每个单词出现的次数
		 //JavaPairRDD<String,Integer> wordsCountRDD = wordsPairRDD.reduceByKey((a,b) -> a+b);
		//相当于 ==>
		JavaPairRDD<String,Integer> wordsCountRDD = wordsPairRDD.reduceByKey(new Function2<Integer, Integer, Integer>() {
			private static final long serialVersionUID = 1L;

			@Override
			public Integer call(Integer v1, Integer v2) throws Exception {
				return v1 + v2;
			}
		});
		 
		 //6.因为只能对key进行排序，所以需要将wordsCountRDD进行key-value倒置，返回新的RDD
		// JavaPairRDD<Integer,String> wordsCountRDD2 = wordsCountRDD.mapToPair(s -> new Tuple2<Integer,String>(s._2, s._1));
		 //相当于 ==>
           JavaPairRDD<Integer,String> wordsCountRDD2 = wordsCountRDD.mapToPair(new PairFunction<Tuple2<String,Integer>, Integer, String>() {
			private static final long serialVersionUID = 1L;
			@Override
			public Tuple2<Integer, String> call(Tuple2<String, Integer> t) throws Exception {
				return new Tuple2<Integer,String>(t._2,t._1);
			}
		});
         
		 //7.对wordsCountRDD2进行排序,降序desc
		 JavaPairRDD<Integer,String> wordsCountRDD3 = wordsCountRDD2.sortByKey(false);
		 
		 //8.只取前10个
		 List<Tuple2<Integer, String>>  result = wordsCountRDD3.take(10);
		 
		 //9.打印
	//	 result.forEach(t -> System.out.println(t._2 + "   " + t._1));
		 for(Tuple2<Integer, String> t : result){
			 System.out.println(t._2 + "   " + t._1);
		 }
		 
		 //10.将结果保存到文件中
		 wordsCountRDD3.saveAsTextFile(args[2]);
		 sc.close();
	 }
}

使用mvn package打包命令