启动的时候，一直要求输入namenode密码：

查看namenode的日志如下：

2019-03-28 18:38:08,961 INFO org.apache.hadoop.ipc.Client: Retrying connect to server: datanode1/192.168.197.136:8485. Already tried 47 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=100, sleepTime=10000 MILLISECONDS)
2019-03-28 18:38:09,208 INFO org.apache.hadoop.ipc.Client: Retrying connect to server: namenode2/192.168.197.135:8030. Already tried 29 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=100, sleepTime=10000 MILLISECONDS)
2019-03-28 18:38:18,958 INFO org.apache.hadoop.ipc.Client: Retrying connect to server: datanode2/192.168.197.137:8485. Already tried 48 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=100, sleepTime=10000 MILLISECONDS)
2019-03-28 18:38:18,961 INFO org.apache.hadoop.ipc.Client: Retrying connect to server: datanode3/192.168.197.138:8485. Already tried 48 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=100, sleepTime=10000 MILLISECONDS)

 怀疑是其他服务器，namenode2/192.168.197.135，datanode1/192.168.197.136，datanode2/192.168.197.137，datanode3/192.168.197.138

上面没有authorized_keys文件中namenode1/192.1682.197.134的公钥不对。

在namenode1/192.1682.197.134中重新执行ssh-keygen -t rsa，生成公钥和私钥，发现私钥真的和其他服务器上存储的namenode1/192.1682.197.134的公钥不匹配。

更改其他服务器上authorized_keys中namenode1/192.1682.197.134的公钥后，重启hdfs正常启动。