在idea上运行Hadoop的MapReduce

项目场景：

在idea上运行Hadoop的MapReduce

---

问题描述：

处理得到的数据不在返回结果文件中，reducer类没有问题，调试发现reducer类未运行：

	IntWritable key2 = new IntWritable();
    Text value2 = new Text();

    protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
/**数据结构：
 * emp:7499,ALLEN,SALESMAN,7698,1981/2/20,1600,300,30
 * dept:20,RESEARCH,DALLAS
 */
        System.out.println("偏移量：" + key + ",value : " + value.toString());
//1、分词
        String[] splits = value.toString().split(",");
//2、区别emp和dept. check if its emp or dept  file
        if (splits.length >= 8){ //读取的是emp表数据
            String empName = splits[1]; // employee name
            String empDept = splits[7]; //deptno
            key2.set(Integer.parseInt(empDept));
            value2.set(empName);
        }else{ //读取是dept表的数据
            String detpNo = splits[0];
// remember to add a flag to identify if its dept file data
            String deptName = "*" + splits[1];//加*的目的是标识当前的数据是属于部门表里面的
            key2.set(Integer.parseInt(detpNo));
            value2.set(deptName);
        }

	}

---

原因分析：

在map类中忘记调用context.write，进而导致没有数据传给reducer。

---

解决方案：

加上

IntWritable key2 = new IntWritable();
    Text value2 = new Text();

    protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
/**数据结构：
 * emp:7499,ALLEN,SALESMAN,7698,1981/2/20,1600,300,30
 * dept:20,RESEARCH,DALLAS
 */
        System.out.println("偏移量：" + key + ",value : " + value.toString());
//1、分词
        String[] splits = value.toString().split(",");
//2、区别emp和dept. check if its emp or dept  file
        if (splits.length >= 8){ //读取的是emp表数据
            String empName = splits[1]; // employee name
            String empDept = splits[7]; //deptno
            key2.set(Integer.parseInt(empDept));
            value2.set(empName);
        }else{ //读取是dept表的数据
            String detpNo = splits[0];
// remember to add a flag to identify if its dept file data
            String deptName = "*" + splits[1];//加*的目的是标识当前的数据是属于部门表里面的
            key2.set(Integer.parseInt(detpNo));
            value2.set(deptName);
        }
//3、通过context写出去. Mapper output
        context.write(key2,value2);
    }