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:
> Hello,
>
> I followed the instructions on the FAQ page to configure and compile openmpi
> so that it should work with Torque
I believe the problem is a missing "#" character. So, the correct way to
specify aTorque/PBS parameter is:
#PBS -N Test
Since you specified "PBS -N Test" you are getting the message "PBS" not
found. "PBS" is being interpreted as a command and not a parameter.
J.W. (Pat) O'Bryant,Jr.
Business
Happy New Year, Gus!
Yes, I'm using affinity. This is my openmpi-mca-params.conf file:
---
# Use RSH instead SSH
pls_rsh_agent=rsh
# Turning on processor affinity
mpi_paffinity_alone=1
# Include using eth1.
btl_tcp_if_include=eth1
# Exclude using lo and eth0.
btl_tcp_if_exclude=lo,eth0
---
I r
Hmm, perhaps not so excellent. It seems to me that openmpi-1.4a1r22335
does have the fixes to trac 2043. So, either the fixes are
insufficient and/or you're experiencing a different problem. I'll see
if I can reproduce your problem, but I'm not confident here.
Louis Rossi wrote:
Hi E
Hi, I have created a cluster with openmpi this way.
1 - Configured SSH with authorized_keys from the server to the nodes.
2 - In */etc/openmpi-default-hostfile\ I have typed the IP of every nodes
and the server.
3 - Afterward I have created a Java application of two threads just to type
a text on
Hi Eugene,
I believe that r22335 did solve resolve the issue. The problem was
between my screen and my chair. Last night, I reset my paths, but the
directory was appended to the paths which had the old mpi directory
information. I think it was linking with the old libraries. I'll try
it
This problem has been solved and yes I accidentally deleted the #'s. Thank
you so much!
Steve
List-Post: users@lists.open-mpi.org
Date: Tue, 5 Jan 2010 06:29:50 -0600
> From: pat.o'bry...@exxonmobil.com
> Subject: Re: [OMPI users] Cannot get OpenMPI 1.3.3 to work with Torque
>2.4.2
> To:
Hi Ilya
1) The only thing that stands out as very different from
what I do here is your configuration flag "--enable-mpi-threads".
Maybe some OpenMPI pro/developer could shed some light about this,
whether that flag could be a potential source for the errors you see.
Considering that when you s
You might want to try something simpler than java to start with. For example:
mpirun -np 4 hostname
(where hostname is the POSIX command line app, not an MPI app)
You should see the hostnames from the first 4 hosts in your hostfile (assuming
each one of them has 1 process slot). Then try runn
On Jan 4, 2010, at 2:02 PM, Eugene Loh wrote:
> Maybe the experts on this list can comment on what *should* be happening
> inside OMPI.
>
> Meanwhile, you should probably avoid MPI_Sendrecv_replace if you care about
> performance. The function is mostly a convenience function and if you care
Sorry If I misunderstood but, Should the mpirun -np4 hostname command return
all the machines no the openmpi-default-hostfile?
My host file has just two ips, the first is the server and the second a
node.
192.168.0.1
192.168.0.3
It just return the name of my server 4 times.
The same happens wit
Hi Leonardo
It depends on how many "slots" (a.k.a. CPUs/cores) you have per node.
You can add the flag "-bynode" to the mpiexec command line
to force it to run on different nodes.
"man mpiexec" is a helpful source of information about these things.
I hope this helps.
Gus Correa
---
12 matches
Mail list logo
