Hadoop实例WordCount程序一步一步运行

虽说现在用Eclipse下开发Hadoop程序很方便了，但是命令行方式对于小程序开发验证很方便。这是初学hadoop时的笔记，记录下来以备查。

　　1. 经典的WordCound程序（WordCount.java），可参见 hadoop0.18文档

import  java.io.IOException;
import  java.util.ArrayList;
import  java.util.Iterator;
import  java.util.List;
import  java.util.StringTokenizer;

import  org.apache.hadoop.conf.Configuration;
import  org.apache.hadoop.conf.Configured;
import  org.apache.hadoop.fs.Path;
import  org.apache.hadoop.io.IntWritable;
import  org.apache.hadoop.io.LongWritable;
import  org.apache.hadoop.io.Text;
import  org.apache.hadoop.mapred.FileInputFormat;
import  org.apache.hadoop.mapred.FileOutputFormat;
import  org.apache.hadoop.mapred.JobClient;
import  org.apache.hadoop.mapred.JobConf;
import  org.apache.hadoop.mapred.MapReduceBase;
import  org.apache.hadoop.mapred.Mapper;
import  org.apache.hadoop.mapred.OutputCollector;
import  org.apache.hadoop.mapred.Reducer;
import  org.apache.hadoop.mapred.Reporter;
import  org.apache.hadoop.util.Tool;
import  org.apache.hadoop.util.ToolRunner;

public   class  WordCount  extends  Configured  implements  Tool {

public   static   class  MapClass  extends  MapReduceBase  implements
Mapper < LongWritable, Text, Text, IntWritable >  {

private   final   static  IntWritable one  =   new  IntWritable( 1 );
private  Text word  =   new  Text();

public   void  map(LongWritable key, Text value,
OutputCollector < Text, IntWritable >  output, Reporter reporter)
throws  IOException {
String line  =  value.toString();
StringTokenizer itr  =   new  StringTokenizer(line);
while  (itr.hasMoreTokens()) {
word.set(itr.nextToken());
output.collect(word, one);
}
}
}

/**
* A reducer class that just emits the sum of the input values.
*/
public   static   class  Reduce  extends  MapReduceBase  implements
Reducer < Text, IntWritable, Text, IntWritable >  {

public   void  reduce(Text key, Iterator < IntWritable >  values,
OutputCollector < Text, IntWritable >  output, Reporter reporter)
throws  IOException {
int  sum  =   0 ;
while  (values.hasNext()) {
sum  +=  values.next().get();
}
output.collect(key,  new  IntWritable(sum));
}
}

static   int  printUsage() {
System.out.println( " wordcount [-m <maps>] [-r <reduces>] <input> <output> " );
ToolRunner.printGenericCommandUsage(System.out);
return   - 1 ;
}

/**
* The main driver for word count map/reduce program. Invoke this method to
* submit the map/reduce job.
*
*  @throws  IOException
* When there is communication problems with the job tracker.
*/
public   int  run(String[] args)  throws  Exception {
JobConf conf  =   new  JobConf(getConf(), WordCount. class );
conf.setJobName( " wordcount " );

//  the keys are words (strings)
conf.setOutputKeyClass(Text. class );
//  the values are counts (ints)
conf.setOutputValueClass(IntWritable. class );

conf.setMapperClass(MapClass. class );
conf.setCombinerClass(Reduce. class );
conf.setReducerClass(Reduce. class );

List < String >  other_args  =   new  ArrayList < String > ();
for  ( int  i  =   0 ; i  <  args.length;  ++ i) {
try  {
if  ( " -m " .equals(args[i])) {
conf.setNumMapTasks(Integer.parseInt(args[ ++ i]));
}  else   if  ( " -r " .equals(args[i])) {
conf.setNumReduceTasks(Integer.parseInt(args[ ++ i]));
}  else  {
other_args.add(args[i]);
}
}  catch  (NumberFormatException except) {
System.out.println( " ERROR: Integer expected instead of  "
+  args[i]);
return  printUsage();
}  catch  (ArrayIndexOutOfBoundsException except) {
System.out.println( " ERROR: Required parameter missing from  "
+  args[i  -   1 ]);
return  printUsage();
}
}

//  Make sure there are exactly 2 parameters left.
if  (other_args.size()  !=   2 ) {
System.out.println( " ERROR: Wrong number of parameters:  "
+  other_args.size()  +   "  instead of 2. " );
return  printUsage();
}
FileInputFormat.setInputPaths(conf, other_args.get( 0 ));
FileOutputFormat.setOutputPath(conf,  new  Path(other_args.get( 1 )));

JobClient.runJob(conf);
return   0 ;
}

public   static   void  main(String[] args)  throws  Exception {
int  res  =  ToolRunner.run( new  Configuration(),  new  WordCount(), args);
System.exit(res);
}

}

　　2. 保证hadoop集群是配置好了的，单机的也好。新建一个目录，比如 /home/admin/WordCount
　　编译WordCount.java程序。

javac  - classpath  / home / admin / hadoop / hadoop - 0.19 . 1 - core.jar WordCount.java  - d  / home / admin / WordCount

　　3. 编译完后在/home/admin/WordCount目录会发现三个class文件 WordCount.class，WordCount$Map.class，WordCount$Reduce.class。
　　cd 进入 /home/admin/WordCount目录,然后执行：

jar cvf WordCount.jar  * . class

　　就会生成 WordCount.jar 文件。

　　4. 构造一些输入数据
　　input1.txt和input2.txt的文件里面是一些单词。如下：

[admin@host WordCount]$ cat input1.txt
Hello, i love china
are you ok ?
[admin@host WordCount]$ cat input2.txt
hello, i love word
You are ok

　　在hadoop上新建目录，和put程序运行所需要的输入文件:

hadoop fs  - mkdir  / tmp / input
hadoop fs  - mkdir  / tmp / output
hadoop fs  - put input1.txt  / tmp / input /
hadoop fs  - put input2.txt  / tmp / input /

　　5. 运行程序,会显示job运行时的一些信息。

[admin@host WordCount]$ hadoop jar WordCount.jar WordCount  / tmp / input  / tmp / output
10 / 09 / 16   22 : 49 : 43  WARN mapred.JobClient: Use GenericOptionsParser  for  parsing the arguments. Applications should implement Tool  for  the same.
10 / 09 / 16   22 : 49 : 43  INFO mapred.FileInputFormat: Total input paths to process : 2
10 / 09 / 16   22 : 49 : 43  INFO mapred.JobClient: Running job: job_201008171228_76165
10 / 09 / 16   22 : 49 : 44  INFO mapred.JobClient: map  0 %  reduce  0 %
10 / 09 / 16   22 : 49 : 47  INFO mapred.JobClient: map  100 %  reduce  0 %
10 / 09 / 16   22 : 49 : 54  INFO mapred.JobClient: map  100 %  reduce  100 %
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Job complete: job_201008171228_76165
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Counters:  16
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: File Systems
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: HDFS bytes read = 62
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: HDFS bytes written = 73
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Local bytes read = 152
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Local bytes written = 366
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Job Counters
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Launched reduce tasks = 1
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Rack - local map tasks = 2
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Launched map tasks = 2
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Map - Reduce Framework
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Reduce input groups = 11
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Combine output records = 14
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Map input records = 4
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Reduce output records = 11
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Map output bytes = 118
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Map input bytes = 62
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Combine input records = 14
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Map output records = 14
10 / 09 / 16   22 : 49 : 55  INFO mapred.JobClient: Reduce input records = 14

　　6. 查看运行结果

[admin@host WordCount]$ hadoop fs  - ls  / tmp / output /
Found  2  items
drwxr - x ---   -  admin admin  0   2010 - 09 - 16   22 : 43   / tmp / output / _logs
- rw - r -----   1  admin admin  102   2010 - 09 - 16   22 : 44   / tmp / output / part - 00000
[admin@host WordCount]$ hadoop fs  - cat  / tmp / output / part - 00000
Hello,  1
You  1
are  2
china  1
hello,  1
i  2
love  2
ok  1
ok ?   1
word  1
you  1