hadoop学习--安装使用-CSDN博客

本文链接：https://blog.csdn.net/ToBeAndNotToBe/article/details/6669008

原理普及：

关于hadoop的原理，可以去很多地方普及，这里不再啰嗦。

hadoop内部结构：

hadoop执行map-reduce流程图：

使用教程：

1.下载hadoop core 源代码：http://apache.etoak.com//hadoop/common/hadoop-0.20.2/hadoop-0.20.2.tar.gz

2.安装jdk（pass），修改hadoop脚本中java home变量：songjings-macpro31:conf songjing$ sudo vim bin/hadoop-env.sh

3.跑一个hadoop自带的单机模式的例子，单词统计：

bin/hadoop jar hadoop-0.20.2-examples.jar wordcount test-in test-out

其中test-in为一个目录，目录中是多个文件，文件内容即为待统计的字符串

test-out为该例子输出结果目录

4.查看计算过程和结果：

songjing$ bin/hadoop jar hadoop-0.20.2-examples.jar wordcount test-in test-out
11/08/08 13:37:46 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
11/08/08 13:37:46 INFO input.FileInputFormat: Total input paths to process : 2 ##输入文件为两个
11/08/08 13:37:47 INFO mapred.JobClient: Running job: job_local_0001 ##启动一个job
11/08/08 13:37:48 INFO input.FileInputFormat: Total input paths to process : 2
11/08/08 13:37:48 INFO mapred.MapTask: io.sort.mb = 100
11/08/08 13:37:50 INFO mapred.MapTask: data buffer = 79691776/99614720
11/08/08 13:37:50 INFO mapred.MapTask: record buffer = 262144/327680
11/08/08 13:37:50 INFO mapred.JobClient: map 0% reduce 0% ##map开始
11/08/08 13:37:50 INFO mapred.MapTask: Starting flush of map output
11/08/08 13:37:50 INFO mapred.MapTask: Finished spill 0
11/08/08 13:37:50 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000000_0 is done. And is in the process of commiting
11/08/08 13:37:50 INFO mapred.LocalJobRunner:
11/08/08 13:37:50 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000000_0' done.
11/08/08 13:37:50 INFO mapred.MapTask: io.sort.mb = 100
11/08/08 13:37:51 INFO mapred.JobClient: map 100% reduce 0%
11/08/08 13:37:51 INFO mapred.MapTask: data buffer = 79691776/99614720
11/08/08 13:37:51 INFO mapred.MapTask: record buffer = 262144/327680
11/08/08 13:37:51 INFO mapred.MapTask: Starting flush of map output
11/08/08 13:37:51 INFO mapred.MapTask: Finished spill 0
11/08/08 13:37:51 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000001_0 is done. And is in the process of commiting
11/08/08 13:37:51 INFO mapred.LocalJobRunner:
11/08/08 13:37:51 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000001_0' done.
11/08/08 13:37:51 INFO mapred.LocalJobRunner:
11/08/08 13:37:51 INFO mapred.Merger: Merging 2 sorted segments ##两个文件的结果进行合并
11/08/08 13:37:51 INFO mapred.Merger: Down to the last merge-pass, with 2 segments left of total size: 76 bytes
11/08/08 13:37:51 INFO mapred.LocalJobRunner:
11/08/08 13:37:51 INFO mapred.TaskRunner: Task:attempt_local_0001_r_000000_0 is done. And is in the process of commiting
11/08/08 13:37:51 INFO mapred.LocalJobRunner:
11/08/08 13:37:51 INFO mapred.TaskRunner: Task attempt_local_0001_r_000000_0 is allowed to commit now
11/08/08 13:37:51 INFO output.FileOutputCommitter: Saved output of task 'attempt_local_0001_r_000000_0' to test-out ##map结束，reduce开始
11/08/08 13:37:51 INFO mapred.LocalJobRunner: reduce > reduce
11/08/08 13:37:51 INFO mapred.TaskRunner: Task 'attempt_local_0001_r_000000_0' done.
11/08/08 13:37:52 INFO mapred.JobClient: map 100% reduce 100%
11/08/08 13:37:52 INFO mapred.JobClient: Job complete: job_local_0001
11/08/08 13:37:52 INFO mapred.JobClient: Counters: 12
11/08/08 13:37:52 INFO mapred.JobClient: FileSystemCounters
11/08/08 13:37:52 INFO mapred.JobClient: FILE_BYTES_READ=467984
11/08/08 13:37:52 INFO mapred.JobClient: FILE_BYTES_WRITTEN=513615
11/08/08 13:37:52 INFO mapred.JobClient: Map-Reduce Framework
11/08/08 13:37:52 INFO mapred.JobClient: Reduce input groups=5
11/08/08 13:37:52 INFO mapred.JobClient: Combine output records=6
11/08/08 13:37:52 INFO mapred.JobClient: Map input records=2
11/08/08 13:37:52 INFO mapred.JobClient: Reduce shuffle bytes=0
11/08/08 13:37:52 INFO mapred.JobClient: Reduce output records=5
11/08/08 13:37:52 INFO mapred.JobClient: Spilled Records=12
11/08/08 13:37:52 INFO mapred.JobClient: Map output bytes=81
11/08/08 13:37:52 INFO mapred.JobClient: Combine input records=8
11/08/08 13:37:52 INFO mapred.JobClient: Map output records=8
11/08/08 13:37:52 INFO mapred.JobClient: Reduce input records=6
songjings-macpro31:hadoop-0.20.2 songjing$ cd test-out/
songjings-macpro31:test-out songjing$ ls

part-r-00000

察看最后的结果：

songjings-macpro31:test-out songjing$ more part-r-00000
by 1
goodbye 1
hadoop 2
hello 2

world 2

5.如何写job：

实现mapper和reducer对应的方法map(),reduce():

map方法：

 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);
      }
    }
  }

以上map方法扫描文件中的每一个字符并将起存储为key-value，其中key为单词，value为1

reduce方法：

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;
      for (IntWritable val : values) {
        sum += val.get();
      }
      result.set(sum);
      context.write(key, result);
    }
  }

reduce工作对每个key对应的list进行统计个数

写完map和reduce方法后，需要写主程序来调用map和reduce方法执行job工作：

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);//设置跑job的map-reduce jar包
    job.setMapperClass(TokenizerMapper.class);//指定map实现类
    job.setCombinerClass(IntSumReducer.class);//指定对map结果按照key合并的类
    job.setReducerClass(IntSumReducer.class);//指定reduce类
    job.setOutputKeyClass(Text.class);//设置输出字段的key为text
    job.setOutputValueClass(IntWritable.class);//设置输出字段的value为int类型
    FileInputFormat.addInputPath(job, new Path(otherArgs[0]));//设置要分析的文件路径
    FileOutputFormat.setOutputPath(job, new Path(otherArgs[1]));//设置输出的文件路径
    System.exit(job.waitForCompletion(true) ? 0 : 1);
  }

注意：output路径不能是已存在目录，不然hadoop会报错：

// Ensure that the output directory is set and not already there
    Path outDir = getOutputPath(job);
    if (outDir == null && job.getNumReduceTasks() != 0) {
      throw new InvalidJobConfException("Output directory not set in JobConf.");
    }
    if (outDir != null) {
      FileSystem fs = outDir.getFileSystem(job);
      // normalize the output directory
      outDir = fs.makeQualified(outDir);
      setOutputPath(job, outDir);
      // check its existence
      if (fs.exists(outDir)) {
        throw new FileAlreadyExistsException("Output directory " + outDir + 
                                             " already exists");
      }
    }