Hadoop-MapReduce-ReduceJoin和MapJoin分析及案例实操-连载中

末蓝、 2023-03-01 11:53 43阅读 0赞

### 文章目录 ###

*   *  Join多种应用
     *   *  1 Reduce Join
         *  2 Reduce Join案例实操
         *  3 Map Join
         *  4 Map Join案例实操

## Join多种应用 ##

### 1 Reduce Join ###

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70]

### 2 Reduce Join案例实操 ###

**1）需求**

订单数据表t\_order（order.txt）

<table> 
 <thead> 
  <tr> 
   <th>id</th> 
   <th><font>pid</font></th> 
   <th>amount</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>1001</td> 
   <td><font>01</font></td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>1002</td> 
   <td><font>02</font></td> 
   <td>2</td> 
  </tr> 
  <tr> 
   <td>1003</td> 
   <td><font>03</font></td> 
   <td>3</td> 
  </tr> 
  <tr> 
   <td>1004</td> 
   <td><font>01</font></td> 
   <td>4</td> 
  </tr> 
  <tr> 
   <td>1005</td> 
   <td><font>02</font></td> 
   <td>5</td> 
  </tr> 
  <tr> 
   <td>1006</td> 
   <td><font>03</font></td> 
   <td>6</td> 
  </tr> 
 </tbody> 
</table>

商品信息表t\_product（product.txt）

<table> 
 <thead> 
  <tr> 
   <th><font>pid</font></th> 
   <th>pname</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td><font>01</font></td> 
   <td>小米</td> 
  </tr> 
  <tr> 
   <td><font>02</font></td> 
   <td>华为</td> 
  </tr> 
  <tr> 
   <td><font>03</font></td> 
   <td>格力</td> 
  </tr> 
 </tbody> 
</table>

将商品信息表中数据根据商品**pid**合并到订单数据表中。

最终数据形式

<table> 
 <thead> 
  <tr> 
   <th>id</th> 
   <th>pname</th> 
   <th>amount</th> 
  </tr> 
 </thead> 
 <tbody> 
  <tr> 
   <td>1001</td> 
   <td>小米</td> 
   <td>1</td> 
  </tr> 
  <tr> 
   <td>1004</td> 
   <td>小米</td> 
   <td>4</td> 
  </tr> 
  <tr> 
   <td>1002</td> 
   <td>华为</td> 
   <td>2</td> 
  </tr> 
  <tr> 
   <td>1005</td> 
   <td>华为</td> 
   <td>5</td> 
  </tr> 
  <tr> 
   <td>1003</td> 
   <td>格力</td> 
   <td>3</td> 
  </tr> 
  <tr> 
   <td>1006</td> 
   <td>格力</td> 
   <td>6</td> 
  </tr> 
 </tbody> 
</table>

**2）需求分析**

通过将关联条件作为 Map 输出的 key，将两表满足 Join 条件的数据并携带数据所来源的文件信息，发往同一个ReduceTask，在Reduce中进行数据的串联。

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 1]  
**3）代码实现**

（1）创建商品和订单合并后的Bean类

public class TableBean implements Writable {
        
        private String id; //订单id
        private String pid; //商品id
        private int amount; //商品数量
        private String pname; //商品名称
        private String flag;  //标志字段,order  pd
    
        public TableBean() {
        
        }
    
        //get和set方法略...
    
        @Override
        public void write(DataOutput out) throws IOException {
        
            out.writeUTF(id);
            out.writeUTF(pid);
            out.writeInt(amount);
            out.writeUTF(pname);
            out.writeUTF(flag);
        }
    
        @Override
        public void readFields(DataInput in) throws IOException {
        
            id = in.readUTF();
            pid = in.readUTF();
            amount = in.readInt();
            pname = in.readUTF();
            flag = in.readUTF();
        }
    
        @Override
        public String toString() {
        
            return id + "\t" + pname + "\t" + amount;
        }
    }

（2）编写TableMapper类

public class TableMapper extends Mapper<LongWritable, Text,Text,TableBean> {
        
    
        private String fileName;
        private Text outK = new Text();
        private TableBean outV = new TableBean();
    
        @Override
        protected void setup(Context context) throws IOException, InterruptedException {
        
            //在setup里面写,是因为一个maptask执行一次,而一个maptask执行的数据刚好就是一个文件
            //通过context对象获取切片信息,然后通过切片信息获取文件名称
            FileSplit fileSplit = (FileSplit) context.getInputSplit();
            fileName = fileSplit.getPath().getName();
        }
    
        @Override
        protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        
            String line = value.toString();
            if(fileName.contains("order")){
          //对order表做操作
                //1001	01	1
                String[] split = line.split("\t");
                //封装outKV
                outK.set(split[1]);
                outV.setId(split[0]);
                outV.setPid(split[1]);
                outV.setAmount(Integer.parseInt(split[2]));
                outV.setPname("");
                outV.setFlag("order");
            }else{
          //对pd表做操作
                //01	小米
                String[] split = line.split("\t");
                //封装outKV
                outK.set(split[0]);
                outV.setId("");
                outV.setPid(split[0]);
                outV.setAmount(0);
                outV.setPname(split[1]);
                outV.setFlag("pd");
            }
            //写出outKV
            context.write(outK,outV);
        }
    }

（3）编写TableReducer类

public class TableReducer extends Reducer<Text, TableBean, TableBean, NullWritable> {
        
        @Override
        protected void reduce(Text key, Iterable<TableBean> values, Context context) throws IOException, InterruptedException {
        
            //用来存储orderbean的数组
            ArrayList<TableBean> orderBeans = new ArrayList<>();
            //用来存pdbean
            TableBean pdBean = new TableBean();
            
            for (TableBean value : values) {
        
                //判断数据来自哪个表
                if("order".equals(value.getFlag())){
         //订单表
                    //创建一个临时TableBean对象接收value
                    //因为Iterable每次迭代都是同一个对象(地址值不变)，只是改变对象元素的值
                    //所以需要创建临时TableBean来接收每次value的值
                    TableBean tempBean = new TableBean();
                    try {
        
                        BeanUtils.copyProperties(tempBean,value);
                    } catch (Exception e) {
        
                        e.printStackTrace();
                    }
                    orderBeans.add(tempBean);
                }else {
         //商品表
                    try {
        
                        BeanUtils.copyProperties(pdBean,value);
                    } catch (Exception e) {
        
                        e.printStackTrace();
                    }
                }
            }
    
            //遍历集合orderBeans,替换掉每个orderBean的pid为pname,然后写出
            for (TableBean orderBean : orderBeans) {
        
                String pname = pdBean.getPname();
                orderBean.setPname(pname);
                //写出修改后的orderBean对象
                context.write(orderBean,NullWritable.get());
            }
    
        }
    }

（4）编写TableDriver类

public class TableDriver {
        
        public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException {
        
            Job job = Job.getInstance(new Configuration());
    
            job.setJarByClass(TableDriver.class);
            job.setMapperClass(TableMapper.class);
            job.setReducerClass(TableReducer.class);
    
            job.setMapOutputKeyClass(Text.class);
            job.setMapOutputValueClass(TableBean.class);
    
            job.setOutputKeyClass(TableBean.class);
            job.setOutputValueClass(NullWritable.class);
    
            FileInputFormat.setInputPaths(job, new Path("D:\\input\\inputtable"));
            FileOutputFormat.setOutputPath(job, new Path("D:\\hadoop\\reducejoin4"));
    
            boolean b = job.waitForCompletion(true);
            System.exit(b ? 0 : 1);
    
        }
    }

**4）总结**  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 2]

### 3 Map Join ###

**1）使用场景**

Map Join适用于一张表十分小、一张表很大的场景。

**2）优点**

思考：在Reduce端处理过多的表，非常容易产生数据倾斜。怎么办？

在Map端缓存多张表，提前处理业务逻辑，这样增加Map端业务，减少Reduce端数据的压力，尽可能的减少数据倾斜。

**3）具体办法：采用DistributedCache**

（1）在Mapper的setup阶段，将文件读取到缓存集合中。

（2）在Driver驱动类中加载缓存。

//缓存普通文件到Task运行节点。
    job.addCacheFile(new URI("file:///e:/cache/pd.txt"));

### 4 Map Join案例实操 ###

**1）需求**

订单数据表t\_order（order.txt）

商品信息表t\_product（product.txt）

将商品信息表中数据根据商品**pid**合并到订单数据表中。

最终数据形式

**2）需求分析**

MapJoin适用于关联表中有小表的情形。  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 3]  
**3）实现代码**

（1）先在MapJoinDriver驱动类中添加缓存文件

public class MapJoinDriver {
        
        public static void main(String[] args) throws Exception {
        
            // 1 获取job信息
            Configuration conf = new Configuration();
            Job job = Job.getInstance(conf);
            // 2 设置加载jar包路径
            job.setJarByClass(MapJoinDriver.class);
            // 3 关联mapper
            job.setMapperClass(MapJoinMapper.class);
            // 4 设置Map输出KV类型
            job.setMapOutputKeyClass(Text.class);
            job.setMapOutputValueClass(NullWritable.class);
            // 5 设置最终输出KV类型
            job.setOutputKeyClass(Text.class);
            job.setOutputValueClass(NullWritable.class);
    
            // 加载缓存数据
            job.addCacheFile(new URI("file:///D:/input/inputtablecache/pd.txt"));
            // Map端Join的逻辑不需要Reduce阶段，设置reduceTask数量为0
            job.setNumReduceTasks(0);
    
            // 6 设置输入输出路径
            FileInputFormat.setInputPaths(job, new Path("D:\\input\\inputtable2"));
            FileOutputFormat.setOutputPath(job, new Path("D:\\hadoop\\mapjoin"));
            // 7 提交
            boolean b = job.waitForCompletion(true);
            System.exit(b ? 0 : 1);
    
        }
    }

（2）在MapJoinMapper类中的setup方法中读取缓存文件

public class MapJoinMapper extends Mapper<LongWritable, Text, Text, NullWritable> {
        
        HashMap<String, String> pdMap = new HashMap();
        private Text outK = new Text();
    
        @Override
        protected void setup(Context context) throws IOException, InterruptedException {
        
            //获取缓存文件  在driver添加的缓存文件
            URI[] cacheFiles = context.getCacheFiles();
            //获取fs对象,是不是要创建对应的文件输入流
            FileSystem fs = FileSystem.get(context.getConfiguration());
            FSDataInputStream fis = fs.open(new Path(cacheFiles[0]));
            //因为hdfs提供的FSDataInputStream不能按行读取,所以要使用转换流包装一下,得到BufferedReader
            BufferedReader reader = new BufferedReader(new InputStreamReader(fis, "UTF-8"));
    
            //按行读取,逐行处理
            String line;
            while (StringUtils.isNotEmpty(line = reader.readLine())) {
        
                //01	小米
                String[] split = line.split("\t");
                pdMap.put(split[0], split[1]);
            }
            //关流
            IOUtils.closeStream(reader);
            fs.close();
        }
    
        @Override
        protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        
            //获取一行数据  1001	01	1
            String orderLine = value.toString();
            String[] split = orderLine.split("\t");
            //通过切割出来的pid,去pdMap里面获取到pname
            String pname = pdMap.get(split[1]);
            //封装outK 其实就是将每一行数据的pid替换成pname  pname是去内存里面的pdMap里面获取到的
            outK.set(split[0] + "\t" + pname + "\t" + split[2]);
            //写出
            context.write(outK,NullWritable.get());
        }
    }

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70]: /images/20230208/d065ec5fa52a4707a143af2e84fc0660.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 1]: /images/20230208/8c736858fee244518e006ab0be792d06.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 2]: /images/20230208/a7ee662045c3405498d3423c4e6b4ce6.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzMyNzI3MDk1_size_16_color_FFFFFF_t_70 3]: /images/20230208/b6d3376dc9a443d0bab4049244546106.png