为什么我的Hadoop映射器类中被读取了两次序列文件？

StackOverflow https://stackoverflow.com/questions/9514710

  •  14-11-2019
  •  | 
  •  

题

我有一个1264条记录的序列文件。每个密钥对每个记录都是唯一的。我的问题是，我的映射似乎是读取这个文件两次，否则它是读取的两次。对于Sanity检查，我已经写了一个小实用程序类来读取序列文件，实际上只有1264条记录（即序列文件.Reader）。

在我的reducer中，我应该只得到1条迭代，但是，当我迭代可迭代（迭代器）时，我每键获得2条记录（总是每个键，而不是每键的其他东西或其他键）。

我的工作的日志记录输出如下。我不确定为什么，但为什么它是“处理的总输入路径”是2？当我运行我的工作时，我尝试了-dmapred.input.dir= / data和-dmapred.input.dir= / data / part-r-00000，但仍然，处理的总路径为2。

任何想法都很欣赏。 

12/03/01 05:28:30 INFO jvm.JvmMetrics: Initializing JVM Metrics with processName=JobTracker, sessionId=
12/03/01 05:28:30 INFO input.FileInputFormat: Total input paths to process : 2
12/03/01 05:28:31 INFO mapred.JobClient: Running job: job_local_0001
12/03/01 05:28:31 INFO input.FileInputFormat: Total input paths to process : 2
12/03/01 05:28:31 INFO mapred.MapTask: io.sort.mb = 100
12/03/01 05:28:31 INFO mapred.MapTask: data buffer = 79691776/99614720
12/03/01 05:28:31 INFO mapred.MapTask: record buffer = 262144/327680
12/03/01 05:28:31 INFO mapred.MapTask: Starting flush of map output
12/03/01 05:28:31 INFO mapred.MapTask: Finished spill 0
12/03/01 05:28:31 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000000_0 is done. And is in the process of commiting
12/03/01 05:28:31 INFO mapred.LocalJobRunner:
12/03/01 05:28:31 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000000_0' done.
12/03/01 05:28:31 INFO mapred.MapTask: io.sort.mb = 100
12/03/01 05:28:31 INFO mapred.MapTask: data buffer = 79691776/99614720
12/03/01 05:28:31 INFO mapred.MapTask: record buffer = 262144/327680
12/03/01 05:28:31 INFO mapred.MapTask: Starting flush of map output
12/03/01 05:28:31 INFO mapred.MapTask: Finished spill 0
12/03/01 05:28:31 INFO mapred.TaskRunner: Task:attempt_local_0001_m_000001_0 is done. And is in the process of commiting
12/03/01 05:28:31 INFO mapred.LocalJobRunner:
12/03/01 05:28:31 INFO mapred.TaskRunner: Task 'attempt_local_0001_m_000001_0' done.
12/03/01 05:28:31 INFO mapred.LocalJobRunner:
12/03/01 05:28:31 INFO mapred.Merger: Merging 2 sorted segments
12/03/01 05:28:31 INFO mapred.Merger: Down to the last merge-pass, with 2 segments left of total size: 307310 bytes
12/03/01 05:28:31 INFO mapred.LocalJobRunner:
12/03/01 05:28:32 INFO mapred.TaskRunner: Task:attempt_local_0001_r_000000_0 is done. And is in the process of commiting
12/03/01 05:28:32 INFO mapred.LocalJobRunner:
12/03/01 05:28:32 INFO mapred.TaskRunner: Task attempt_local_0001_r_000000_0 is allowed to commit now
12/03/01 05:28:32 INFO mapred.JobClient:  map 100% reduce 0%
12/03/01 05:28:32 INFO output.FileOutputCommitter: Saved output of task 'attempt_local_0001_r_000000_0' to results
12/03/01 05:28:32 INFO mapred.LocalJobRunner: reduce > reduce
12/03/01 05:28:32 INFO mapred.TaskRunner: Task 'attempt_local_0001_r_000000_0' done.
12/03/01 05:28:33 INFO mapred.JobClient:  map 100% reduce 100%
12/03/01 05:28:33 INFO mapred.JobClient: Job complete: job_local_0001
12/03/01 05:28:33 INFO mapred.JobClient: Counters: 12
12/03/01 05:28:33 INFO mapred.JobClient:   FileSystemCounters
12/03/01 05:28:33 INFO mapred.JobClient:     FILE_BYTES_READ=1320214
12/03/01 05:28:33 INFO mapred.JobClient:     FILE_BYTES_WRITTEN=1275041
12/03/01 05:28:33 INFO mapred.JobClient:   Map-Reduce Framework
12/03/01 05:28:33 INFO mapred.JobClient:     Reduce input groups=1264
12/03/01 05:28:33 INFO mapred.JobClient:     Combine output records=0
12/03/01 05:28:33 INFO mapred.JobClient:     Map input records=2528
12/03/01 05:28:33 INFO mapred.JobClient:     Reduce shuffle bytes=0
12/03/01 05:28:33 INFO mapred.JobClient:     Reduce output records=2528
12/03/01 05:28:33 INFO mapred.JobClient:     Spilled Records=5056
12/03/01 05:28:33 INFO mapred.JobClient:     Map output bytes=301472
12/03/01 05:28:33 INFO mapred.JobClient:     Combine input records=0
12/03/01 05:28:33 INFO mapred.JobClient:     Map output records=2528
12/03/01 05:28:33 INFO mapred.JobClient:     Reduce input records=2528
.

我的地图per类很简单。它在文本文件中读取。对于每行，它将“M”附加到线路。 

public class MyMapper extends Mapper<LongWritable, Text, LongWritable, Text> {

 private static final Log _log = LogFactory.getLog(MyMapper.class);

 @Override
 public void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
  String s = (new StringBuilder()).append(value.toString()).append("m").toString();
  context.write(key, new Text(s));
  _log.debug(key.toString() + " => " + s);
 }
}
.

我的减速机课也很简单。它只是将“r”附加到行。 

public class MyReducer extends Reducer<LongWritable, Text, LongWritable, Text> {

private static final Log _log = LogFactory.getLog(MyReducer.class);

@Override
public void reduce(LongWritable key, Iterable<Text> values, Context context) throws IOException, InterruptedException {
 for(Iterator<Text> it = values.iterator(); it.hasNext();) {
  Text txt = it.next();
  String s = (new StringBuilder()).append(txt.toString()).append("r").toString();
  context.write(key, new Text(s));
  _log.debug(key.toString() + " => " + s);
  }
 }
}
.

我的工作类如下。 

public class MyJob extends Configured implements Tool {

public static void main(String[] args) throws Exception {
 ToolRunner.run(new Configuration(), new MyJob(), args);
}

@Override
public int run(String[] args) throws Exception {
 Configuration conf = getConf();
 Path input = new Path(conf.get("mapred.input.dir"));
 Path output = new Path(conf.get("mapred.output.dir"));

 System.out.println("input = " + input);
 System.out.println("output = " + output);

 Job job = new Job(conf, "dummy job");
 job.setMapOutputKeyClass(LongWritable.class);
 job.setMapOutputValueClass(Text.class);
 job.setOutputKeyClass(LongWritable.class);
 job.setOutputValueClass(Text.class);

 job.setMapperClass(MyMapper.class);
 job.setReducerClass(MyReducer.class);

 FileInputFormat.addInputPath(job, input);
 FileOutputFormat.setOutputPath(job, output);

 job.setJarByClass(MyJob.class);

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

我的输入数据看起来如下。 

T, T
T, T
T, T
F, F
F, F
F, F
F, F
T, F
F, T
.

运行我的作业后，我得到了以下的输出。 

0   T, Tmr
0   T, Tmr
6   T, Tmr
6   T, Tmr
12  T, Tmr
12  T, Tmr
18  F, Fmr
18  F, Fmr
24  F, Fmr
24  F, Fmr
30  F, Fmr
30  F, Fmr
36  F, Fmr
36  F, Fmr
42  T, Fmr
42  T, Fmr
48  F, Tmr
48  F, Tmr
.

我做了什么问题了吗？我尝试了以下方式来运行我的工作，并且在这种方法中，该文件只能读取一次。为什么是这样？ system.out.println（InPath）和System.out.println（Sutpath）值是相同的！帮助？ 

public class MyJob2 {

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: MyJob2 <in> <out>");
      System.exit(2);
    }

    String sInput = args[0];
    String sOutput = args[1];

    Path input = new Path(sInput);
    Path output = new Path(sOutput);

    System.out.println("input = " + input);
    System.out.println("output = " + output);

    Job job = new Job(conf, "dummy job");
    job.setMapOutputKeyClass(LongWritable.class);
    job.setMapOutputValueClass(Text.class);
    job.setOutputKeyClass(LongWritable.class);
    job.setOutputValueClass(Text.class);

    job.setMapperClass(MyMapper.class);
    job.setReducerClass(MyReducer.class);

    FileInputFormat.addInputPath(job, input);
    FileOutputFormat.setOutputPath(job, output);

    job.setJarByClass(MyJob2.class);

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

有帮助吗？

解决方案

i got help from the hadoop mailing list. my problem was with the line below.

FileInputFormat.addInputPath(job, input);

this line simply appends input back to config. after commenting this line out, the input file is read only once now. in fact, i also commented out the other line,

FileOutputFormat.setOutputPath(job, output);

and everything still works.

其他提示

I've had a similar problem, but for a different reason: linux apparently created a hidden copy of my input file (~input.txt), so that's a second way of getting this error..

许可以下： CC-BY-SA 和 归因
不隶属于 StackOverflow
scroll top