MapReduce分区和排序

Posted areyouready

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了MapReduce分区和排序相关的知识,希望对你有一定的参考价值。

一、排序

排序:

需求:根据用户每月使用的流量按照使用的流量多少排序

接口-->WritableCompareable

    排序操作在hadoop中属于默认的行为。默认按照字典殊勋排序。
    
排序的分类:

    1)部分排序
    
    2)全排序
    
    3)辅助排序
    
    4)二次排序
    
Combiner 合并

    父类Reducer
    局部汇总 ,减少网络传输量 ,进而优化程序。
    
    注意:求平均值?
    
    3  5  7  2  6
    
    mapper: (3 + 5 + 7)/3 = 5
            (2 + 6)/2 = 4
            
    reducer:(5+4)/2
    
    只能应用在不影响最终业务逻辑的情况下

二、分区和排序实例

1.Mapper类

package com.css.flowsort;

import java.io.IOException;

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

public class FlowSortMapper extends Mapper<LongWritable, Text, FlowBean, Text>{

    @Override
    protected void map(LongWritable key, Text value, Context context)
            throws IOException, InterruptedException {
        // 1.获取一行数据
        String line = value.toString();
        // 2.切割
        String[] fields = line.split("	");
        // 3.取出关键字段
        long upFlow = Long.parseLong(fields[1]);
        long dfFlow = Long.parseLong(fields[2]);
        // 4.写出到reducer阶段
        context.write(new FlowBean(upFlow, dfFlow), new Text(fields[0]));
    }
}

2.Reducer类

package com.css.flowsort;

import java.io.IOException;

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

public class FlowSortReducer extends Reducer<FlowBean, Text, Text, FlowBean>{

    @Override
    protected void reduce(FlowBean key, Iterable<Text> value, Context context)
            throws IOException, InterruptedException {     
        context.write(value.iterator().next(), key);
    }
}

3.封装类

package com.css.flowsort;

import java.io.DataInput;
import java.io.DataOutput;
import java.io.IOException;

import org.apache.hadoop.io.WritableComparable;

// 封装类 直接完成排序
public class FlowBean implements WritableComparable<FlowBean> {

    // 定义属性
    private long upFlow;
    private long dfFlow;
    private long flowSum;
    
    // 无参构造
    public FlowBean() {        
    }
    
    // 有参构造
    public FlowBean(long upFlow,long dfFlow){
        this.upFlow = upFlow;
        this.dfFlow = dfFlow;
        this.flowSum = upFlow + dfFlow;
    }
    
    public long getUpFlow() {
        return upFlow;
    }

    public void setUpFlow(long upFlow) {
        this.upFlow = upFlow;
    }

    public long getDfFlow() {
        return dfFlow;
    }

    public void setDfFlow(long dfFlow) {
        this.dfFlow = dfFlow;
    }

    public long getFlowSum() {
        return flowSum;
    }

    public void setFlowSum(long flowSum) {
        this.flowSum = flowSum;
    }
    
    // 反序列化
    @Override
    public void readFields(DataInput in) throws IOException {
        upFlow = in.readLong();
        dfFlow = in.readLong();
        flowSum = in.readLong();
    }

    // 序列化
    @Override
    public void write(DataOutput out) throws IOException {
        out.writeLong(upFlow);
        out.writeLong(dfFlow);
        out.writeLong(flowSum);
    }
    
    @Override
    public String toString() {
        return upFlow + "	" + dfFlow + "	" + flowSum;
    }

    // 排序
    @Override
    public int compareTo(FlowBean o) {
        // 倒序
        return this.flowSum > o.getFlowSum() ? -1 : 1;
    }
}

4.自定义分区类

package com.css.flowsort;

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

public class FlowSortPartitioner extends Partitioner<FlowBean, Text>{

    // 根据手机号前三位进行分区
    @Override
    public int getPartition(FlowBean key, Text value, int numPartitions) {
        // 获取手机号前三位
        String phoneNum = value.toString().substring(0, 3);
        // 分区
        int partitioner = 4;
        if ("135".equals(phoneNum)) {
            return 0;
        }else if ("137".equals(phoneNum)) {
            return 1;
        }else if ("138".equals(phoneNum)) {
            return 2;
        }else if ("139".equals(phoneNum)) {
            return 3;
        }
        return partitioner;
    }
}

5.Driver类

package com.css.flowsort;

import java.io.IOException;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
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 FlowSortDriver {
    public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException {
        // 1.获取job信息
        Configuration conf = new Configuration();
        Job job = Job.getInstance(conf);
        
        // 2.获取jar包
        job.setJarByClass(FlowSortDriver.class);
        
        // 3.获取自定义的mapper与reducer类
        job.setMapperClass(FlowSortMapper.class);
        job.setReducerClass(FlowSortReducer.class);
        
        // 4.设置map输出的数据类型
        job.setMapOutputKeyClass(FlowBean.class);
        job.setMapOutputValueClass(Text.class);
        
        // 5.设置reduce输出的数据类型(最终的数据类型)
        job.setOutputKeyClass(Text.class);
        job.setOutputValueClass(FlowBean.class);
        
        //添加自定义分区
        job.setPartitionerClass(FlowSortPartitioner.class);
        job.setNumReduceTasks(5);
        
        // 6.设置输入存在的路径与处理后的结果路径
        FileInputFormat.setInputPaths(job, new Path("c:/flow1024/in"));
        FileOutputFormat.setOutputPath(job, new Path("c:/flow1024/out1"));
        
        // 7.提交任务
        boolean rs = job.waitForCompletion(true);
        System.out.println(rs ? 0 : 1);
    }
}

6.输入的文件part-r-00000

13480253104    120    1320    1440
13502468823    735    11349    12084
13510439658    1116    954    2070
13560436326    1136    94    1230
13560436666    1136    94    1230
13560439658    918    4938    5856
13602846565    198    910    1108
13660577991    660    690    1350
13719199419    240    0    240
13726130503    299    681    980
13726238888    2481    24681    27162
13760778710    120    120    240
13822544101    264    0    264
13884138413    4116    1432    5548
13922314466    3008    3720    6728
13925057413    11058    4243    15301
13926251106    240    0    240
13926435656    132    1512    1644
15013685858    369    338    707
15889002119    938    380    1318
15920133257    316    296    612
18212575961    1527    2106    3633
18320173382    9531    212    9743

7.如果第5步Driver类中的红色部分去掉,则输出全局排序后的文件part-r-00000

13726238888    2481    24681    27162
13925057413    11058    4243    15301
13502468823    735    11349    12084
18320173382    9531    212    9743
13922314466    3008    3720    6728
13560439658    918    4938    5856
13884138413    4116    1432    5548
18212575961    1527    2106    3633
13510439658    1116    954    2070
13926435656    132    1512    1644
13480253104    120    1320    1440
13660577991    660    690    1350
15889002119    938    380    1318
13560436326    1136    94    1230
13560436666    1136    94    1230
13602846565    198    910    1108
13726130503    299    681    980
15013685858    369    338    707
15920133257    316    296    612
13822544101    264    0    264
13760778710    120    120    240
13719199419    240    0    240
13926251106    240    0    240

8.如果第5步Driver类中的红色部分不去掉,则输出分区加排序后的文件

(1)part-r-00000
13502468823    735    11349    12084
13560439658    918    4938    5856
13510439658    1116    954    2070
13560436666    1136    94    1230
13560436326    1136    94    12302)part-r-00001
13726238888    2481    24681    27162
13726130503    299    681    980
13760778710    120    120    240
13719199419    240    0    2403)part-r-00002
13884138413    4116    1432    5548
13822544101    264    0    2644)part-r-00003
13925057413    11058    4243    15301
13922314466    3008    3720    6728
13926435656    132    1512    1644
13926251106    240    0    2405)part-r-00004
18320173382    9531    212    9743
18212575961    1527    2106    3633
13480253104    120    1320    1440
13660577991    660    690    1350
15889002119    938    380    1318
13602846565    198    910    1108
15013685858    369    338    707
15920133257    316    296    612

 


以上是关于MapReduce分区和排序的主要内容,如果未能解决你的问题,请参考以下文章

Hadoop Mapreduce分区分组二次排序

我们真的需要在 MapReduce 框架中进行排序吗?

MapReduce 学习4 ---- 自定义分区自定义排序自定义组分

大数据之Hadoop(MapReduce):WritableComparable排序案例实操(区内排序)

Spark的shuffle和MapReduce的shuffle对比

MapReduce中shuffle过程