在 Hadoop 中实现自定义可写? [英] Implementation of custom Writable in Hadoop?

问题描述

我在 Hadoop 中定义了一个自定义的 Writable 类，但是 Hadoop 在运行我的程序时给了我以下错误消息.

I have defined a custom Writable class in Hadoop, but Hadoop gives me the following error message when running my program.

java.lang.RuntimeException: java.lang.NullPointerException
at org.apache.hadoop.util.ReflectionUtils.newInstance(ReflectionUtils.java:115)
at org.apache.hadoop.io.SortedMapWritable.readFields(SortedMapWritable.java:180)
at EquivalenceClsAggValue.readFields(EquivalenceClsAggValue.java:82)
at org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:67)
at org.apache.hadoop.io.serializer.WritableSerialization$WritableDeserializer.deserialize(WritableSerialization.java:40)
at org.apache.hadoop.mapred.Task$ValuesIterator.readNextValue(Task.java:1282)
at org.apache.hadoop.mapred.Task$ValuesIterator.next(Task.java:1222)
at org.apache.hadoop.mapred.Task$CombineValuesIterator.next(Task.java:1301)
at Mondrian$Combine.reduce(Mondrian.java:119)
at Mondrian$Combine.reduce(Mondrian.java:1)
at org.apache.hadoop.mapred.Task$OldCombinerRunner.combine(Task.java:1442)
at org.apache.hadoop.mapred.MapTask$MapOutputBuffer.sortAndSpill(MapTask.java:1436)
at org.apache.hadoop.mapred.MapTask$MapOutputBuffer.flush(MapTask.java:1298)
at org.apache.hadoop.mapred.MapTask.runOldMapper(MapTask.java:437)
at org.apache.hadoop.mapred.MapTask.run(MapTask.java:372)
at org.apache.hadoop.mapred.Child$4.run(Child.java:255)
at java.security.AccessController.doPrivileged(Native Method)
at javax.security.auth.Subject.doAs(Subject.java:415)
at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1136)
at org.apache.hadoop.mapred.Child.main(Child.java:249)

原因:java.lang.NullPointerException在 java.util.concurrent.ConcurrentHashMap.hash(ConcurrentHashMap.java:332)....

Caused by: java.lang.NullPointerException at java.util.concurrent.ConcurrentHashMap.hash(ConcurrentHashMap.java:332)....

EquivalenceClsAggValue 是我定义的 Writable 类的名称，这是我的类:

EquivalenceClsAggValue is the name of the Writable class I've defined and this is my class:

public class EquivalenceClsAggValue implements WritableComparable<EquivalenceClsAggValue>{

public ArrayList<SortedMapWritable> aggValues;  
public EquivalenceClsAggValue(){        

    aggValues = new ArrayList<SortedMapWritable>();
}
@Override
public void readFields(DataInput arg0) throws IOException {

    int size = arg0.readInt();

    for (int i=0;i<size;i++){
        SortedMapWritable tmp = new SortedMapWritable();
        tmp.readFields(arg0);
        aggValues.add(tmp);
    }       
}

@Override
public void write(DataOutput arg0) throws IOException {

    //write the size first
    arg0.write(aggValues.size());

    //write each element
    for (SortedMapWritable s:aggValues){
        s.write(arg0);
    }

}

我想知道问题的根源是什么.

I wonder to know what is the source of the problem.

推荐答案

看起来你的 write(DataOutput) 方法有错误:

Looks like an error in your write(DataOutput) method:

@Override
public void write(DataOutput arg0) throws IOException {
  //write the size first
  // arg0.write(aggValues.size()); // here you're writing an int as a byte

  // try this instead:
  arg0.writeInt(aggValues.size()); // actually write int as an int

  //..

查看 DataOutput.write(int) 与 DataOutput.writeInt(int)

我还将修改您在 readFields 中创建的 SortedMapWritable tmp 局部变量以使用 ReflectionUtils.newInstance():

I'd also amend your creation of the SortedMapWritable tmp local variable in readFields to use ReflectionUtils.newInstance():

@Override
public void readFields(DataInput arg0) throws IOException {

  int size = arg0.readInt();

  for (int i=0;i<size;i++){
    SortedMapWritable tmp = ReflectionUtils.newInstance(
        SortedMapWritable.class, getConf());
    tmp.readFields(arg0);
    aggValues.add(tmp);
  }       
}

请注意，要使其正常工作，您还需要修改类签名以扩展 Configurable (这样 Hadoop 会在您的对象最初创建时注入 Configuration 对象):

Note for this to work, you'll also need to amend you class signature to extend Configurable (such that Hadoop will inject a Configuration object when your object is initially created):

public class EquivalenceClsAggValue 
          extends Configured 
          implements WritableComparable<EquivalenceClsAggValue> {

这篇关于在 Hadoop 中实现自定义可写?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！