so wrong mailing list huh?<div><br></div><div>sorry to bother</div><div><br></div><div>thanks for your time<br><br><div class="gmail_quote">On Wed, Nov 30, 2011 at 1:52 PM, Lloyd Brown <span dir="ltr"><<a href="mailto:lloyd_brown@byu.edu">lloyd_brown@byu.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Ricardo,<br>
<br>
Have you seen section 4.1.4 ("Mapping a Queue to a Subset of Resources")<br>
in the Torque documentation? It might give you some ideas. However,<br>
the short answer to your question, as seen in that section is this:<br>
<br>
> TORQUE does not currently provide a simple mechanism for mapping queues to nodes. However, schedulers such as Moab and Maui can provide this functionality.<br>
<br>
<br>
Lloyd Brown<br>
Systems Administrator<br>
Fulton Supercomputing Lab<br>
Brigham Young University<br>
<a href="http://marylou.byu.edu" target="_blank">http://marylou.byu.edu</a><br>
<div><div class="h5"><br>
<br>
<br>
On 11/30/2011 12:37 PM, Ricardo Román Brenes wrote:<br>
> Hello everyone thanks for the time of reading and the long post :P<br>
><br>
><br>
> The question is about multiple queues with Torque:<br>
><br>
><br>
> We have here different clusternodes with difrente architectures:<br>
> 4 PS-3<br>
> 3 CPU+GPU<br>
> 2 CPU<br>
><br>
> and i want to be able to send jobs to each of hte nodes independly<br>
> (using torque). Im guessing that having several queues and that each<br>
> node belonging to a queue in particular and then submittint jobs to that<br>
> queue will do the trick:<br>
><br>
> say i got 4 queues<br>
> IBMCELL with the 4 PS-3<br>
> TESLA with the 3 nodes that have GPUs<br>
> XEON with te 5 nodes that have Xeons (which in turn 3 of them have<br>
> teslas :P)<br>
><br>
> and when i submit a job:<br>
> qsub -q IBMCELL a.pbs<br>
> should run on the PS-3 only, but im not being able to make it work like<br>
> that.<br>
><br>
> As a test i made 2 queues in the PS3 pbs_server ("uno" and "dos"):<br>
><br>
> #<br>
> # Create queues and set their attributes.<br>
> #<br>
> #<br>
> # Create and define queue uno<br>
> #<br>
</div></div>> *create queue uno<br>
> **set queue uno queue_type = Execution<br>
> **set queue uno acl_host_enable = False<br>
> **set queue uno acl_hosts = zarate-0+zarate-1<br>
> **set queue uno enabled = True<br>
> **set queue uno started = True<br>
> *#<br>
<div class="im">> # Create and define queue dos<br>
> #<br>
</div>> *create queue dos<br>
> **set queue dos queue_type = Execution<br>
> **set queue dos acl_host_enable = **False**<br>
> **set queue dos acl_hosts = zarate-2+zarate-3<br>
> **set queue dos enabled = True<br>
> **set queue dos started = True<br>
> *#<br>
<div class="im">> # Set server attributes.<br>
> #<br>
> set server scheduling = True<br>
> set server acl_hosts = zarate-0<br>
> set server log_events = 511<br>
> set server mail_from = adm<br>
> set server scheduler_iteration = 600<br>
> set server node_check_rate = 150<br>
> set server tcp_timeout = 6<br>
> set server next_job_number = 22<br>
><br>
><br>
</div>> and i changed the _nodes_ file in the server_priv directory so it is<br>
<div><div class="h5">> like this (zarate are just the hostname :P):<br>
><br>
><br>
> zarate-0 np=2 uno<br>
> zarate-1 np=2 uno<br>
> zarate-2 np=2 dos<br>
> zarate-3 np=2 dos<br>
><br>
><br>
><br>
> but its not working... when i launch a job:<br>
><br>
> #PBS -N mpi_hello<br>
> /usr/local/bin/mpiexec -n 8 /home/rroman/a.out<br>
><br>
><br>
> with teh command:<br>
><br>
> #PBS -N mpi_hello<br>
><br>
> /usr/local/bin/mpiexec -n 8 /home/rroman/a.out<br>
><br>
><br>
> the output file is:<br>
><br>
> zarate-1: hello world from process 2 of 8<br>
> zarate-2: hello world from process 5 of 8<br>
> zarate-2: hello world from process 6 of 8<br>
> zarate-3: hello world from process 0 of 8<br>
> zarate-3: hello world from process 7 of 8<br>
> zarate-1: hello world from process 3 of 8<br>
> zarate-0: hello world from process 4 of 8<br>
> zarate-3: hello world from process 1 of 8<br>
><br>
><br>
><br>
> And there it shows that the job is running in ALL the nodes instead of<br>
> running only in zarate-0 and zarate-1 as the queue said (according to me :P)<br>
><br>
><br>
><br>
><br>
> SO! the question is: is it possible to do waht i want like this? and if<br>
> so, what am i doing wrong! :P<br>
><br>
> Thank you Kay!<br>
><br>
> -ricardo<br>
><br>
><br>
><br>
</div></div>> _______________________________________________<br>
> torqueusers mailing list<br>
> <a href="mailto:torqueusers@supercluster.org">torqueusers@supercluster.org</a><br>
> <a href="http://www.supercluster.org/mailman/listinfo/torqueusers" target="_blank">http://www.supercluster.org/mailman/listinfo/torqueusers</a><br>
_______________________________________________<br>
torqueusers mailing list<br>
<a href="mailto:torqueusers@supercluster.org">torqueusers@supercluster.org</a><br>
<a href="http://www.supercluster.org/mailman/listinfo/torqueusers" target="_blank">http://www.supercluster.org/mailman/listinfo/torqueusers</a><br>
</blockquote></div><br></div>