环境

Linux：Hadoop2.x

Windows：jdk1.8、Maven3、IDEA2021

步骤

编程分析

编程分析包括：
1.数据过程分析：数据从输入到输出的过程分析。
2.数据类型分析：Map的输入输出类型，Reduce的输入输出类型；
编程分析决定了我们该如何编写代码。

新建Maven工程

打开IDEA–>点击File–>New–>Project

选择Maven–>点击Next

选择一个空目录作为项目目录，目录名称例如：wordcount，建议目录路径不包含中文和空格，点击Finish

添加依赖

修改pom.xml，添加如下依赖

    <dependencies><dependency><groupId>org.apache.hadoop</groupId><artifactId>hadoop-common</artifactId><version>2.7.3</version></dependency><dependency><groupId>org.apache.hadoop</groupId><artifactId>hadoop-client</artifactId><version>2.7.3</version></dependency><dependency><groupId>org.apache.hadoop</groupId><artifactId>hadoop-hdfs</artifactId><version>2.7.3</version></dependency><dependency><groupId>org.apache.hadoop</groupId><artifactId>hadoop-mapreduce-client-core</artifactId><version>2.7.3</version></dependency></dependencies>

加载依赖

新建包

在src\main\java目录下，新建包:org.example

填入org.example，效果如下：

新建类

在org.example包下，新建出三个类，分别为:MyMapper、MyReducer、MyMain，效果如下：

编写Map程序

编辑MyMapper类，步骤如下：

1.继承Mapper
2.重写map()方法
3.编写Map逻辑代码：1.v1由Text类型转换为String2.按空格进行分词：split(" ")方法3.输出k2, v2

编写Reduce程序

编辑MyReducer类，步骤如下：

1.继承Reducer
2.重写reduce()方法
3.编写Reduce逻辑代码：1.k4 = k32.v4 = v3元素的和3.输出k4, v4

编写Main程序（Driver程序）

编辑MyMain类，步骤如下：

1. 创建一个job和任务入口(指定主类)
2. 指定job的mapper和输出的类型<k2 v2>
3. 指定job的reducer和输出的类型<k4  v4>
4. 指定job的输入和输出路径
5. 执行job

思考

代码编写完成后，可以先在Windows本地运行吗？

打包

看到BUILD SUCCESS为打包成功

打包后得到的jar包，在项目的target目录下

提交到Hadoop集群运行

1.将上一步打包得到的jar包，上传到linux

2.启动hadoop集群

start-all.sh

3.运行jar包

从Linux本地上传一个文件到hdfs

hdfs dfs -put 1.txt /input/1.txt

hdfs查看输入数据

运行jar包

hadoop jar wordcount-1.0-SNAPSHOT.jar org.example.MyMain /input/1.txt /output/wordcount

正常运行过程输出如下：

[hadoop@node1 ~]$ hadoop jar wordcount-1.0-SNAPSHOT.jar org.example.MyMain /input/1.txt /output/wordcount
22/03/29 00:23:59 INFO client.RMProxy: Connecting to ResourceManager at node1/192.168.193.140:8032
22/03/29 00:23:59 WARN mapreduce.JobResourceUploader: Hadoop command-line option parsing not performed. Implement the Tool interface and execute your application with ToolRunner to remedy this.
22/03/29 00:24:00 INFO input.FileInputFormat: Total input paths to process : 1
22/03/29 00:24:00 INFO mapreduce.JobSubmitter: number of splits:1
22/03/29 00:24:01 INFO mapreduce.JobSubmitter: Submitting tokens for job: job_1648484275192_0001
22/03/29 00:24:01 INFO impl.YarnClientImpl: Submitted application application_1648484275192_0001
22/03/29 00:24:01 INFO mapreduce.Job: The url to track the job: http://node1:8088/proxy/application_1648484275192_0001/
22/03/29 00:24:01 INFO mapreduce.Job: Running job: job_1648484275192_0001
22/03/29 00:24:08 INFO mapreduce.Job: Job job_1648484275192_0001 running in uber mode : false
22/03/29 00:24:08 INFO mapreduce.Job:  map 0% reduce 0%
22/03/29 00:24:12 INFO mapreduce.Job:  map 100% reduce 0%
22/03/29 00:24:17 INFO mapreduce.Job:  map 100% reduce 100%
22/03/29 00:24:19 INFO mapreduce.Job: Job job_1648484275192_0001 completed successfully
22/03/29 00:24:19 INFO mapreduce.Job: Counters: 49File System CountersFILE: Number of bytes read=55FILE: Number of bytes written=237261FILE: Number of read operations=0FILE: Number of large read operations=0FILE: Number of write operations=0HDFS: Number of bytes read=119HDFS: Number of bytes written=25HDFS: Number of read operations=6HDFS: Number of large read operations=0HDFS: Number of write operations=2Job Counters Launched map tasks=1Launched reduce tasks=1Data-local map tasks=1Total time spent by all maps in occupied slots (ms)=2290Total time spent by all reduces in occupied slots (ms)=2516Total time spent by all map tasks (ms)=2290Total time spent by all reduce tasks (ms)=2516Total vcore-milliseconds taken by all map tasks=2290Total vcore-milliseconds taken by all reduce tasks=2516Total megabyte-milliseconds taken by all map tasks=2344960Total megabyte-milliseconds taken by all reduce tasks=2576384Map-Reduce FrameworkMap input records=2Map output records=4Map output bytes=41Map output materialized bytes=55Input split bytes=94Combine input records=0Combine output records=0Reduce input groups=3Reduce shuffle bytes=55Reduce input records=4Reduce output records=3Spilled Records=8Shuffled Maps =1Failed Shuffles=0Merged Map outputs=1GC time elapsed (ms)=103CPU time spent (ms)=1200Physical memory (bytes) snapshot=425283584Virtual memory (bytes) snapshot=4223356928Total committed heap usage (bytes)=277348352Shuffle ErrorsBAD_ID=0CONNECTION=0IO_ERROR=0WRONG_LENGTH=0WRONG_MAP=0WRONG_REDUCE=0File Input Format Counters Bytes Read=25File Output Format Counters Bytes Written=25
[hadoop@node1 ~]$ 

查看输出结果

思考

• 如果运行过程报如下错误，该如何解决？
  

• 代码还可以优化吗？如何优化？

完成！enjoy it!