问题
I have a hadoop installation on the Amazon Elastic MapReduce , whenever I try to restart the cluster I get the following error:
/stop-all.sh
no jobtracker to stop
The authenticity of host 'localhost (::1)' can't be established. RSA key fingerprint is
Are you sure you want to continue connecting (yes/no)? yes
localhost: Warning: Permanently added 'localhost' (RSA) to the list of known hosts.
localhost: Permission denied (publickey).
no namenode to stop
localhost: Permission denied (publickey).
localhost: Permission denied (publickey).
Any idea on how to restart hadoop?
回答1:
Following hack worked for me.
I have replaced "ssh" command in sbin/slaves.sh & sbin/hadoop-daemon.sh with "ssh -i ~/.ssh/keyname"
回答2:
I'm using hadoop version 2.4 and this worked for me:
export HADOOP_SSH_OPTS="-i /home/hadoop/mykey.pem"
回答3:
For the stop-all.sh script to work, you probably need to have the same user in all the machines as the user with which you are executing the stop-all.sh script.
Moreover, it seems you do not have a password less ssh setup from the machine you are executing stop-all.sh to rest of the machines that will spare you from manually entering the password for each machine separately. Passwords might be different for the same user for different machines, please don't forget that.
来源：https://stackoverflow.com/questions/18813958/how-to-restart-hadoop-cluster-on-emr