Jeff, Thank you for your advice. By bad. I took the wrong shot, because I tested so many different settings. After I came back to the original network settings, "permission denied', of course disappeared, but the other messages were still there. The master node has two NICs, one for WAN (via another server) with zone=external and the other for the slave node, zone = internal. The NICs on the master are in different subnet. NIC on the slave node is set to 'internal'.Their status was confirmed by firewall-cmd --get-active-zones.
I temporary stopped firewalld and the error messages disappeared. I saw six processes were running on each node, but now the all processes keep running forever with 100% CPU usage. -----Original Message----- From: Jeff Squyres (jsquyres) <jsquy...@cisco.com> To: Open MPI User's List <us...@open-mpi.org> Sent: Mon, May 23, 2016 9:13 am Subject: Re: [OMPI users] problem about mpirun on two nodes On May 21, 2016, at 11:31 PM, dour...@aol.com wrote: > > I encountered a problem about mpirun and SSH when using OMPI 1.10.0 compiled > with gcc, running on centos7.2. > When I execute mpirun on my 2 node cluster, I get the following errors pasted > below. > > [douraku@master home]$ mpirun -np 12 a.out > Permission denied (publickey,gssapi-keyex,gssapi-with-mic). This is the key right here: you got a permission denied error when you (assumedly) tried to execute on the remote server. Triple check your ssh settings to ensure that you can run on the remote server(s) without a password or interactive passphrase entry. -- Jeff Squyres jsquy...@cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/ _______________________________________________ users mailing list us...@open-mpi.org Subscription: https://www.open-mpi.org/mailman/listinfo.cgi/users Link to this post: http://www.open-mpi.org/community/lists/users/2016/05/29282.php
