ssh 升级导致的hadoop 主备切换失败

zkfc的日志信息:

2020-07-27 09:41:16,304 WARN org.apache.hadoop.ha.SshFenceByTcpPort: PATH=$PATH:/sbin:/usr/sbin fuser -v -k -n tcp 9000 via ssh: bash: fuser: command not found
2020-07-27 09:41:16,305 INFO org.apache.hadoop.ha.SshFenceByTcpPort: rc: 127
2020-07-27 09:41:16,305 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Disconnecting from zy03-cap-21m3aser-074 port 22
2020-07-27 09:41:16,305 WARN org.apache.hadoop.ha.NodeFencer: Fencing method org.apache.hadoop.ha.SshFenceByTcpPort(null) was unsuccessful.
2020-07-27 09:41:16,305 ERROR org.apache.hadoop.ha.NodeFencer: Unable to fence service by any configured method.
2020-07-27 09:41:16,305 INFO org.apache.hadoop.ha.SshFenceByTcpPort.jsch: Caught an exception, leaving main loop due to Socket closed
2020-07-27 09:41:16,305 WARN org.apache.hadoop.ha.ActiveStandbyElector: Exception handling the winning of election

配置文件里面配置的fence方法

    <property>
        <name>dfs.ha.fencing.methods</name>
        <value>
                sshfence
                 shell(/bin/true)
        </value>
    </property>

从日志里面看出来,两种fence都失败了
方式一:使用fuser 命令
从日志里面看出来,fuser 命令没有找到,导致失败
解决办法

 yum install -y psmisc

方式二: 使用shell的方式
Fencing method org.apache.hadoop.ha.SshFenceByTcpPort(null) was unsuccessful. 可以看出来,秘钥失效

posted @ 2020-07-27 10:10  北漂-boy  阅读(778)  评论(0编辑  收藏  举报