I suggest you contact the Torque user list about this - it is a Torque configuration issue, not something to do with OMPI.

On Jan 3, 2010, at 10:49 PM, chih lee wrote:


I followed the instructions on the FAQ page to configure and compile openmpi so that it should work with Torque.
./configure --with-tm=/usr/local --prefix=/usr/local
The option --disable-server was used to configure torque on the compute nodes.
I got openmpi compiled without any error message on the head and compute nodes.

I can use
$ mpirun -np 2 --host node1,node2 a.out
to run parallel programs without any problem.

However,  when I submit the following script with qsub

PBS -N Test
PBS -o /home2/user2/test.sh.o
PBS -l nodes=8
mpirun /home2/user2/a.out  # a.out simply prints out # of procs and its ID

I got the following output and error messages.

N. of procs = 1, proc ID = 0

Error messages:
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 3: PBS: not found
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 4: PBS: not found
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 5: PBS: not found
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 6: PBS: not found
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 7: PBS: not found
/var/spool/torque/mom_priv/jobs/198.my_head_node.SC: 8: PBS: not found

I'm new to OpenMPI and Torque. I really appreciate it if you can give me some insights. Thanks!


users mailing list