为简单的hadoop mapreduce作业运行两个映射器和两个reducer

Question

I just wanted to have a better understanding on using multiple mappers and reducers.I want to try this out using a simple hadoop mapreduce Word count job.I want to run two mapper and two reducer for this wordcount job.Is there that I need to configure manually on the configuration files or is it just enough to just make changes on the WordCount.java file.

I'm running this job on a Single node.And I'm running this job as

$ hadoop jar job.jar input output

And i've started

$ hadoop namenode -format
$ hadoop namenode

$ hadoop datanode

sbin$ ./yarn-daemon.sh start resourcemanager sbin$ ./yarn-daemon.sh start resourcemanager

I'm running hadoop-2.0.0-cdh4.0.0

And my WordCount.java file is

package org.apache.hadoop.examples;

import java.io.IOException;
import java.util.StringTokenizer;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.io.IntWritable;
import org.rg.apache.hadoop.fs.Path;
import oapache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.GenericOptionsParser;

public class WordCount {
private static final Log LOG = LogFactory.getLog(WordCount.class);

  public static class TokenizerMapper
       extends Mapper<Object, Text, Text, IntWritable>{

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

    public void map(Object key, Text value, Context context
                    ) throws IOException, InterruptedException {
      StringTokenizer itr = new StringTokenizer(value.toString());
      while (itr.hasMoreTokens()) {
        word.set(itr.nextToken());
        context.write(word, one);
      }
    }
  }

  public static class IntSumReducer
       extends Reducer<Text,IntWritable,Text,IntWritable> {
    private IntWritable result = new IntWritable();

    public void reduce(Text key, Iterable<IntWritable> values,
                       Context context
                       ) throws IOException, InterruptedException {
      int sum = 0;
      //printKeyAndValues(key, values);

      for (IntWritable val : values) {
        sum += val.get();
      LOG.info("val = " + val.get());
      }
      LOG.info("sum = " + sum + " key = " + key);
      result.set(sum);
      context.write(key, result);
      //System.err.println(String.format("[reduce] word: (%s), count: (%d)", key, result.get()));
    }


  // a little method to print debug output
    private void printKeyAndValues(Text key, Iterable<IntWritable> values)
    {
      StringBuilder sb = new StringBuilder();
      for (IntWritable val : values)
      {
        sb.append(val.get() + ", ");
      }
      System.err.println(String.format("[reduce] key: (%s), value: (%s)", key, sb.toString()));
    }
  }

  public static void main(String[] args) throws Exception {
    Configuration conf = new Configuration();
    String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();
    if (otherArgs.length != 2) {
      System.err.println("Usage: wordcount <in> <out>");
      System.exit(2);
    }
    Job job = new Job(conf, "word count");
    job.setJarByClass(WordCount.class);
    job.setMapperClass(TokenizerMapper.class);
    job.setCombinerClass(IntSumReducer.class);
    job.setReducerClass(IntSumReducer.class);
    job.setOutputKeyClass(Text.class);
    job.setOutputValueClass(IntWritable.class);
    FileInputFormat.addInputPath(job, new Path(otherArgs[0]));
    FileOutputFormat.setOutputPath(job, new Path(otherArgs[1]));

    System.exit(job.waitForCompletion(true) ? 0 : 1);
}
}

Could anyone of you help me now to run two mapper and the reducers for this Word count job?

Answer 1

Gladnick: In case you are planning to use the default TextInputFormat , there would be atleast as many mappers at the number of input files (or more depending on the file size). So just put 2 files into your input directories so that you can get 2 mappers running. (Advising this solution, because you plan to run this as a test case).

Now that you have asked for 2 reducers, all you need to do is job.setNumReduceTasks(2) in your main befor submiting the job.

After that just prepare a jar of your application and run that in hadoop pseudo cluster .

In case you need to specify which word to go to which reducer, you can specify that in the Partitioner class.

            Configuration configuration = new Configuration();
        // create a configuration object that provides access to various
        // configuration parameters
        Job job = new Job(configuration, "Wordcount-Vowels & Consonants");
        // create the job object and set job name as Wordcount-Vowels &
        // Consonants
        job.setJarByClass(WordCount.class);
        // set the main class
        job.setNumReduceTasks(2);
        // set the number of reduce tasks required
        job.setMapperClass(WordCountMapper.class);
        // set the map class for the job
        job.setCombinerClass(WordCountCombiner.class);
        // set the combiner class for the job
        job.setPartitionerClass(VowelConsonantPartitioner.class);
        // set the partitioner class for the job
        job.setReducerClass(WordCountReducer.class);
        // set the reduce class for the job
        job.setOutputKeyClass(Text.class);
        // set the output type of key (the word) expected from the job, Text
        // analogous to String
        job.setOutputValueClass(IntWritable.class);
        // set the output type of value (the count) expected from the job,
        // IntWritable analogous to int
        FileInputFormat.addInputPath(job, new Path(args[0]));
        // set the input directory for fetching the input files
        FileOutputFormat.setOutputPath(job, new Path(args[1])); 

This should be the structure of your main program. You may include the combiner and the partitioner in case needed.

Answer 2

For mappers set

mapred.max.split.size 

to half the size of your file.

For reducers set them to 2 explicitly as

 mapred.reduce.tasks=2