hadoop在序列化的时候,如果被序列化的BEAN里面含有其他对象如list等,一定要在readFields方法里面new出这个新对象。要不然由于序列化里数据重用的原因,会导致list被循环利用,从而使多个BEAN对象的list里面的元素叠加到一个BEAN里。
例如:
?
?
class="java">public class UserPortrait2RedisBean
implements Writable
{
private String deviceId;
private String model;
private List<String> tags;
public UserPortrait2RedisBean()
{
this.tags = new ArrayList();
}
public void readFields(DataInput in)
throws IOException
{
this.deviceId = in.readUTF();
this.model = in.readUTF();
int len = in.readInt();
this.tags = new ArrayList();
for (int i = 0; i < len; i++)
this.tags.add(in.readUTF());
}
public void write(DataOutput out)
throws IOException
{
out.writeUTF(this.deviceId);
out.writeUTF(this.model);
out.writeInt(this.tags.size());
for (String tag : this.tags)
out.writeUTF(tag);
}
}
?
?