我在linum机器中使用单节点集群hadoop-2.7.0。我的wordcount作业代码在1个reducer中运行良好。但如果我增加减速器,效果就不好了。它显示以下错误:
15/05/25 21:15:10 INFO util.NativeCodeLoader: Loaded the native-hadoop library
15/05/25 21:15:10 WARN mapred.JobClient: Use GenericOptionsParser for parsing the arguments. Applications should implement Tool for the same.
15/05/25 21:15:10 WARN mapred.JobClient: No job jar file set. User classes may not be found. See JobConf(Class) or JobConf#setJar(String).
15/05/25 21:15:10 WARN snappy.LoadSnappy: Snappy native library is available
15/05/25 21:15:10 INFO snappy.LoadSnappy: Snappy native library loaded
15/05/25 21:15:10 INFO mapred.FileInputFormat: Total input paths to process : 1
15/05/25 21:15:10 INFO mapred.JobClient: Running job: job_local_0001
15/05/25 21:15:11 INFO util.ProcessTree: setsid exited with exit code 0
15/05/25 21:15:11 INFO mapred.Task: Using ResourceCalculatorPlugin : org.apache.hadoop.util.LinuxResourceCalculatorPlugin@5f1fd699
15/05/25 21:15:11 INFO mapred.MapTask: numReduceTasks: 1
15/05/25 21:15:11 INFO mapred.MapTask: io.sort.mb = 100
15/05/25 21:15:11 INFO mapred.MapTask: data buffer = 79691776/99614720
15/05/25 21:15:11 INFO mapred.MapTask: record buffer = 262144/327680
15/05/25 21:15:11 WARN mapred.LocalJobRunner: job_local_0001
java.io.IOException: Illegal partition for am (1)
at org.apache.hadoop.mapred.MapTask$MapOutputBuffer.collect(MapTask.java:1073)
at org.apache.hadoop.mapred.MapTask$OldOutputCollector.collect(MapTask.java:592)
at WordMapper.map(WordMapper.java:24)
at WordMapper.map(WordMapper.java:1)
at org.apache.hadoop.mapred.MapRunner.run(MapRunner.java:50)
at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:436)
at org.apache.hadoop.mapred.MapTask.run(MapTask.java:372)
at org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java:212)
我的getpartition方法如下所示:
public int getPartition(Text key, IntWritable value, int numRedTasks) {
String s = key.toString();
if(s.length() == 1)
{
return 0;
}
else if(s.length() == 2)
{
return 1;
}
else if(s.length() == 3)
{
return 2;
}
else
return 3;
}
在wordcount.class文件中运行方法:
if(input.length < 2)
{
System.out.println("Please provide valid input");
return -1;
}
else
{
JobConf config = new JobConf();
FileInputFormat.setInputPaths(config, new Path(input[0]));
FileOutputFormat.setOutputPath(config, new Path(input[1]));
config.setMapperClass(WordMapper.class);
config.setReducerClass(WordReducer.class);
config.setNumReduceTasks(4);
config.setPartitionerClass(MyPartitioner.class);
config.setMapOutputKeyClass(Text.class);
config.setMapOutputValueClass(IntWritable.class);
config.setOutputKeyClass(Text.class);
config.setOutputValueClass(IntWritable.class);
JobClient.runJob(config);
}
return 0;
}
我的mapper和reducer代码很好,因为带有1个reducer的wordcount作业运行良好。有人能弄明白吗?
2条答案
按热度按时间vngu2lb81#
这可能是由于清管器在操作中由于高默认值而失败,因此可以在其中设置并行。
谢谢,谢里什。
ux6nzvsh2#
您需要在驱动程序类中使用toorunner,并在主类中调用toolrunner。您可以通过将combiner用作工作流的一部分来实现这一点。下面是驱动程序类代码:正如您从下面的代码中看到的,除了mapper和reducer调用之外,还有一个combiner调用。主运行程序中的退出代码是“int exitcode=toolrunner.run(new configuration(),new wordcountwithcombiner(),args);”它在运行时调用tool runner,您可以通过在运行wordcount程序时使用“-d”选项指定要使用的缩减器或Map器的数量。示例命令行类似于“-d mapred.reduce.tasks=2 input-output”