马士兵hadoop第五课：java开发Map/Reduce（转）

马士兵hadoop第一课：虚拟机搭建和安装hadoop及启动

马士兵hadoop第二课：hdfs集群集中管理和hadoop文件操作

马士兵hadoop第三课：java开发hdfs

马士兵hadoop第四课：Yarn和Map/Reduce配置启动和原理讲解

马士兵hadoop第五课：java开发Map/Reduce

配置系统环境变量HADOOP_HOME，指向hadoop安装目录（如果你不想招惹不必要的麻烦，不要在目录中包含空格或者中文字符）
把HADOOP_HOME/bin加到PATH环境变量（非必要，只是为了方便）
如果是在windows下开发，需要添加windows的库文件
把盘中共享的bin目录覆盖HADOOP_HOME/bin
如果还是不行，把其中的hadoop.dll复制到c:\windows\system32目录下，可能需要重启机器
建立新项目，引入hadoop需要的jar文件

代码WordMapper：

import java.io.IOException;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.LongWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Mapper;

public class WordMapper extends Mapper<LongWritable,Text, Text, IntWritable> {

@Override</br>
</span><span style="color: #0000ff">protected</span> <span style="color: #0000ff">void</span> map(LongWritable key, Text value, Mapper&lt;LongWritable, Text, Text, IntWritable&gt;<span style="color: #000000">.Context context)</br>
        </span><span style="color: #0000ff">throws</span><span style="color: #000000"> IOException, InterruptedException {</br>
    String line </span>=<span style="color: #000000"> value.toString();</br>
    String[] words </span>= line.split(" "<span style="color: #000000">);</br>
    </span><span style="color: #0000ff">for</span><span style="color: #000000">(String word : words) {</br>
        context.write(</span><span style="color: #0000ff">new</span> Text(word), <span style="color: #0000ff">new</span> IntWritable(1<span style="color: #000000">));
    }</br>
}</br>

}

代码WordReducer:

import java.io.IOException;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.LongWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Reducer;

public class WordReducer extends Reducer<Text, IntWritable, Text, LongWritable> {
@Override</br>
</span><span style="color: #0000ff">protected</span> <span style="color: #0000ff">void</span> reduce(Text key, Iterable&lt;IntWritable&gt;<span style="color: #000000"> values,</br>
        Reducer</span>&lt;Text, IntWritable, Text, LongWritable&gt;.Context context) <span style="color: #0000ff">throws</span><span style="color: #000000"> IOException, InterruptedException {</br>
    </span><span style="color: #0000ff">long</span> count = 0<span style="color: #000000">;</br>
    </span><span style="color: #0000ff">for</span><span style="color: #000000">(IntWritable v : values) {</br>
        count </span>+=<span style="color: #000000"> v.get();</br>
    }</br>
    context.write(key, </span><span style="color: #0000ff">new</span><span style="color: #000000"> LongWritable(count));</br>
}</br>

}

代码Test：

import org.apache.hadoop.conf.Configuration;

import org.apache.hadoop.fs.Path;

import org.apache.hadoop.io.IntWritable;

import org.apache.hadoop.io.LongWritable;

import org.apache.hadoop.io.Text;

import org.apache.hadoop.mapreduce.Job;

import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;

import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;


public class Test {


public static void main(String[] args) throws Exception {


Configuration conf = new Configuration();

    Job job </span>=<span style="color: #000000"> Job.getInstance(conf);</br></br>

    job.setMapperClass(WordMapper.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br>
    job.setReducerClass(WordReducer.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br>
    job.setMapOutputKeyClass(Text.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br>
    job.setMapOutputValueClass(IntWritable.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br>
    job.setOutputKeyClass(Text.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br>
    job.setOutputValueClass(LongWritable.</span><span style="color: #0000ff">class</span><span style="color: #000000">);</br></br>

    FileInputFormat.setInputPaths(job, </span>"c:/bigdata/hadoop/test/test.txt"<span style="color: #000000">);</br>
    FileOutputFormat.setOutputPath(job, </span><span style="color: #0000ff">new</span> Path("c:/bigdata/hadoop/test/out/"<span style="color: #000000">));</br></br>

    job.waitForCompletion(</span><span style="color: #0000ff">true</span><span style="color: #000000">);</br>
}</br>

}

把hdfs中的文件拉到本地来运行

FileInputFormat.setInputPaths(job, "hdfs://master:9000/wcinput/");
FileOutputFormat.setOutputPath(job, new Path("hdfs://master:9000/wcoutput2/"));

注意这里是把hdfs文件拉到本地来运行，如果观察输出的话会观察到jobID带有local字样
同时这样的运行方式是不需要yarn的(自己停掉yarn服务做实验)
在远程服务器执行

conf.set("fs.defaultFS", "hdfs://master:9000/");

conf.set("mapreduce.job.jar", "target/wc.jar");


conf.set("mapreduce.framework.name", "yarn");


conf.set("yarn.resourcemanager.hostname", "master");


conf.set("mapreduce.app-submission.cross-platform", "true");

FileInputFormat.setInputPaths(job, "/wcinput/");


FileOutputFormat.setOutputPath(job, new Path("/wcoutput3/"));

如果遇到权限问题，配置执行时的虚拟机参数-DHADOOP_USER_NAME=root
也可以将hadoop的四个配置文件拿下来放到src根目录下，就不需要进行手工配置了，默认到classpath目录寻找
或者将配置文件放到别的地方，使用conf.addResource(.class.getClassLoader.getResourceAsStream)方式添加，不推荐使用绝对路径的方式

原文地址：http://www.cnblogs.com/yucongblog/p/6650872.html