Hi Jeff, Thank you for picking up on the thread. I followed your suggestion and I also did some research. I think I fixed the issue. My problem was two things:
1) From the mater node I made sure i could do a password-less ssh connection to the other nodes via; IP address of the nodes, hostname of the nodes and, Fully Qualified Name of the nodes. I noticed that the ssh session with the IP address was fine but I needed to allow/trust the connection by entering “yes” when i first ran password-less ssh sessions by hostname and FQN hostname. That’s interesting to me because all the post, forum, tutorial i read only talk about password-less session with IP addresses and never mentioned by hostname and FQN hostname. Now i wonder is it because in my /etc/hosts file includes IP's and hostnames as well as FQN hostnames ?? 2) I have forgotten to set the environment path in the /etc/environment file. I had to add the following (in bold) PATH="/usr/local/.openmpi/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/usr/games:/usr/local/games" LD_LIBRARY_PATH="/usr/local/.openmpi/lib/" Now i can run; mpirun -hostfile hostsfile -np 72 /media/homes/ompi/./openmpi_hello I can see the 12 cores per node (6 nodes in total) working!! Thank you Jeff (and Gille) for pointing me into the right direction. Now my next step is to bind Python to OpenMPI. Cheers, Eric _____________________________________________________________________________________________________ Eric F. Alemany System Administrator for Research IRT Division of Radiation & Cancer Biology Department of Radiation Oncology Stanford University School of Medicine Stanford, California 94305 Tel:1-650-498-7969<tel:1-650-498-7969> No Texting Fax:1-650-723-7382<tel:1-650-723-7382>
_______________________________________________ users mailing list users@lists.open-mpi.org https://lists.open-mpi.org/mailman/listinfo/users
