两个NameNode都处在StandBy模式

今天发现两个NameNode都处在StandBy模式

尝试重启hdfs，两个NameNode依然处在StandBy模式

后来发现停止HDFS时，NameNode1不能停止

[root@bigdata01-test hadoop]# stop-dfs.sh
Stopping namenodes on [bigdata01-test bigdata02-test]
bigdata01-test: no namenode to stop
bigdata02-test: stopping namenode
bigdata02-test: no datanode to stop
bigdata01-test: no datanode to stop
bigdata04-test: stopping datanode
bigdata05-test: no datanode to stop
bigdata03-test: no datanode to stop
Stopping journal nodes [bigdata01-test bigdata02-test bigdata03-test bigdata04-test bigdata05-test]
bigdata01-test: no journalnode to stop
bigdata04-test: stopping journalnode
bigdata05-test: no journalnode to stop
bigdata02-test: no journalnode to stop
bigdata03-test: no journalnode to stop
Stopping ZK Failover Controllers on NN hosts [bigdata01-test bigdata02-test]
bigdata01-test: stopping zkfc
bigdata02-test: no zkfc to stop          
[root@bigdata01-test hadoop]# jps
8001 QuorumPeerMain
14565 Jps
28333 NameNode
20845 RunJar
[root@bigdata01-test hadoop]#

可能NameNode进程僵死，直接干掉

[root@bigdata01-test hadoop]# kill 28333
[root@bigdata01-test hadoop]# jps
8001 QuorumPeerMain
15063 Jps
20845 RunJar
[root@bigdata01-test hadoop]# 

然后再重启HDFS，问题临时解决了。

再次查看NameNode状态

[root@bigdata01-test ~]# hdfs haadmin -getServiceState nn1
active
[root@bigdata01-test ~]# hdfs haadmin -getServiceState nn2
standby
[root@bigdata01-test ~]#