MR之自定义分区

---

前言

对于MR而言，数据输出的时候是HashPartitioner分区器来进行数据的分区输出，对于这类的分区器，在某些业务情况下不满足要求，这个时候就需要自定义分区器来满足需求。

---

一、如何自定义分区？

需求自定义分区，则需要定义一个class类，需要继承Partitioner类，
重写getPartition放法，再getPartition方法里面实现业务需求。

二、业务需求

1、有如下电话号码的txt文本

13512312341
13612312342
13712312343
13812312344
13612341235
13712341236
13812341237
13823315644
13623345635
13723345636
13823345637

2、需求：需要将相同号码段的手机号码放到同一个文件中，比如135开头的一个文件，136开头的文件

二、自定义分区

1.CusPartition类

package com.hadoop.mapreduce.customPartition;

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

/**
 * @author LengQing
 * @date 2020/5/3 - 13:06
 */
public class CusPartition extends Partitioner<Text, LongWritable> {
    @Override
    public int getPartition(Text key, LongWritable value, int numPartitions) {
        int partition = 3;
        // 获取key的前三位 135 -》0 136 -》1 137 -》2 138 -》 3
        String index = key.toString().substring(0, 3);
        if ("135".equals(index)){
            partition = 0 ;
        } else if ("136".equals(index)){
            partition = 1;
        } else if ("137".equals(index)){
            partition = 2;
        }
        return partition;
    }
}

2.CusParMapper类

package com.hadoop.mapreduce.customPartition;

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

import java.io.IOException;

/**
 * @author LengQing
 * @date 2020/5/3 - 13:05
 */
public class CusParMapper extends Mapper<LongWritable, Text, Text, LongWritable> {
    private Text outputKey = new Text();
    private LongWritable outputValue = new LongWritable(1);
    @Override
    protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        // 数据字段：电话号码 用户名称 用户地址
        String[] values = value.toString().split(" ");
        this.outputKey.set(values[0]);
        context.write(this.outputKey,this.outputValue);
    }
}

3.CusParReduce类

package com.hadoop.mapreduce.customPartition;

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

/**
 * @author LengQing
 * @date 2020/5/3 - 13:05
 */
public class CusParReduce extends Reducer<Text, LongWritable, Text, LongWritable> {
    private LongWritable outputValue = new LongWritable();
    @Override
    protected void reduce(Text key, Iterable<LongWritable> values, Context context) throws IOException, InterruptedException {
        int sum = 0;
        for (LongWritable value : values) {
            sum += value.get();
        }
        this.outputValue.set(sum);
        context.write(key, this.outputValue);
    }
}

4.CusParDriver类

package com.hadoop.mapreduce.customPartition;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
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;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;

/**
 * @author LengQing
 * @date 2020/5/3 - 13:04
 * 需求：按照电话号码将不同电话数据分到不同的文件中
 * 例如：135开头 文件1
 *      136开头 文件2
 *      137开头 文件3
 *      138开头 文件4
 */
public class CusParDriver extends Configured implements Tool {
    private Configuration conf = new Configuration();
    @Override
    public int run(String[] args) throws Exception {

        // 1 实例化Job
        Job job = Job.getInstance(conf, "cusPartition");
        job.setJarByClass(CusParDriver.class);

        //TODO 1、Input阶段 数据格式：[90,(hadoop hive spark....)]
        Path inputPath = new Path(args[0]);
        FileInputFormat.setInputPaths(job, inputPath);

        //TODO 2、map阶段 数据格式：[(hadoop，1),(hive1),(spark,1)....]
        job.setMapperClass(CusParMapper.class);
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(LongWritable.class);

        //TODO 3、shuffle 数据格式：[(hadoop,[1,1]),(hive,[1,1])....]
        job.setPartitionerClass(CusPartition.class);
        //job.setSortComparatorClass(null);
        job.setNumReduceTasks(4);

        //TODO 4、reduce阶段 数据格式：[(hadoop,2),(hive,2)....]
        job.setReducerClass(CusParReduce.class);
        job.setOutputKeyClass(Text.class);
        job.setOutputValueClass(LongWritable.class);

        //TODO 5、output阶段 数据格式：[(hadoop,2),(hive,2)....]
        Path outputPath = new Path(args[1]);
        FileOutputFormat.setOutputPath(job, outputPath);

        return job.waitForCompletion(true) ? 0 : 1;
    }

    public static void main(String[] args) {
        try {
            int status = ToolRunner.run(new CusParDriver(), args);
            System.exit(status);
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

四、注意点

自定义分区后,我对用的输出了4个分区，135、136、137、138四个分区，对用的会输出4个文件，在设置分区数的时候需要大于等于输出的分区数，如果小于，那么对应得程序会报错，大于则多出得会以空文件补充。

---