Error:map.xml could only be replicated to 0 nodes instead of minReplication (=1)

1.场景: hive执行一条sql语句,然后map,reduce做完了,写数据over了

2.错误:
java.lang.RuntimeException: Error caching map.xml: org.apache.hadoop.ipc.RemoteException(java.io.IOException): File /tmp/hive/root/675fa67b-59c3-4de7-99a9-3f5be9f6685c
/hive_2016-07-19_16-11-06_796_8611429774960723114-1/-mr-10013/f1b1cc9d-19ef-448c-8b9d-3e48129c7fec/map.xml could only be replicated to 0 nodes instead of minReplication (=1).
There are 3 datanode(s) running and no node(s) are excluded in this operation.

3.分析:
1>.检查集群服务状态是否正常
2>.新建个txt文件,然后put到hdfs系统,再次检查
3>.磁盘空间不足

4.解决方法:
a.检查各个节点的磁盘是否99%，100%
df -h

b.查询出大于100M的文件
find / -type f -size +100000000c -exec du -sh {} \;

c.删选不要的文件或者日志,删除,降低磁盘使用

rm -f 文件目录

5.验证: