MapReduce的高级特性（4、合并: Combiner）以及案例 distict去重以及MR的核心shuffle（洗牌）

（*）什么是合并？在Map端先进行一次Reducer的操作，Combiner是一种特殊的Reducer

（*）好处：减少Map输出到Reducer中的数据量，从而提高性能

（*）举例：使用Combiner重写WordCount程序

（*）注意事项： ----> 编程：求平均值

     谨慎使用！！！！

（1）并不是所有的问题都可以使用Combiner: 求平均值

（2）引入了Combiner后，不能改变原来的逻辑；如果改变了，想个办法，让他不改变。

Error: java.io.IOException: wrong value class: class org.apache.hadoop.io.DoubleWritable is not class org.apache.hadoop.io.LongWritable

==================================================================

package demo.wc.combiner;


import java.io.IOException;


import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;


//                                           k1            v1     k2        v2
public class WordCountMapper extends Mapper<LongWritable, Text, Text, LongWritable>{


@Override
protected void map(LongWritable k1, Text v1, Context context)
throws IOException, InterruptedException {
/*
* context代表Map的上下文
* 上文：HDFS
* 下文是：Reducer
*/
//数据: I love Beijing
String data = v1.toString();

//分词
String[] words = data.split(" ");

//输出:  k2     v2
for(String w:words){
context.write(new Text(w), new LongWritable(1));
}
}


}

-------------------------------------------------------------------------------------------------------

package demo.wc.combiner;


import java.io.IOException;


import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Reducer;


//                                              k3     v3          k4       v4
public class WordCountReducer extends Reducer<Text, LongWritable, Text, LongWritable> {


@Override
protected void reduce(Text k3, Iterable<LongWritable> v3,Context context) throws IOException, InterruptedException {
/*
* context 代表Reducer上下文
* 上文：mapper
* 下文：HDFS
*/
long total = 0;
for(LongWritable l:v3){
total = total + l.get();
}

//输出  k4     v4
context.write(k3, new LongWritable(total));
}
}

------------------------------------------------------------------------------------------------------

package demo.wc.combiner;


import java.io.IOException;


import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
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 WordCountMain {


public static void main(String[] args) throws Exception {
//创建一个job = mapper + reducer
Job job = Job.getInstance(new Configuration());
//指定job的入口
job.setJarByClass(WordCountMain.class);

//指定任务的mapper和输出数据类型
job.setMapperClass(WordCountMapper.class);
job.setMapOutputKeyClass(Text.class); //指定k2的类型
job.setMapOutputValueClass(LongWritable.class);//指定v2的数据类型

//设置任务的Combiner
job.setCombinerClass(WordCountReducer.class);

//指定任务的reducer和输出数据类型
job.setReducerClass(WordCountReducer.class);
job.setOutputKeyClass(Text.class);//指定k4的类型
job.setOutputValueClass(LongWritable.class);//指定v4的类型

//指定输入的路径和输出的路径
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));

//执行任务
job.waitForCompletion(true);
}
}
=============================================================

MR的编程案例

1、去掉重复数据（distinct关键字）：利用k2和k3的关系

       SQL:  select distinct deptno, job from emp;

   distinct 作用于后面所有的列

package demo.distinct;


import java.io.IOException;


import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;


public class DistinctMapper extends Mapper<LongWritable, Text, Text, NullWritable> {


@Override
protected void map(LongWritable key1, Text value1, Context context)
throws IOException, InterruptedException {

String str = value1.toString().trim();
context.write(new Text(str), NullWritable.get());
}


}

------------------------------------------------------------------------------------------

package demo.distinct;

import java.io.IOException;

import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Reducer;


public class DistinctReducer extends Reducer<Text, NullWritable, Text, NullWritable> {


@Override
protected void reduce(Text k3, Iterable<NullWritable> v3,Context context) throws IOException, InterruptedException {
context.write(k3, NullWritable.get());
}


}

--------------------------------------------------------------------------------------------------

package demo.distinct;


import java.io.IOException;


import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.DoubleWritable;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.NullWritable;
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 DistinctMain {


public static void main(String[] args) throws Exception {
//创建一个job = mapper + reducer
Job job = Job.getInstance(new Configuration());
//指定job的入口
job.setJarByClass(DistinctMain.class);

//指定任务的mapper和输出数据类型
job.setMapperClass(DistinctMapper.class);
job.setMapOutputKeyClass(Text.class); //指定k2的类型
job.setMapOutputValueClass(NullWritable.class);//指定v2的数据类型


//指定任务的reducer和输出数据类型
job.setReducerClass(DistinctReducer.class);
job.setOutputKeyClass(Text.class);//指定k4的类型
job.setOutputValueClass(NullWritable.class);//指定v4的类型

//指定输入的路径和输出的路径
FileInputFormat.setInputPaths(job, new Path(args[0]));
FileOutputFormat.setOutputPath(job, new Path(args[1]));

//执行任务
job.waitForCompletion(true);
}


}