<font face="verdana,sans-serif">Thanks Bryant for details about tpn, i can not find any details about tpn in torque manual.<br>using tpn, i can start tasks on different node, but still it does not meet the expected behaviour.<br>
<br>1. If my jobs scripts looks like this<br> #PBS -l nodes=4,tpn=4<br> mpirun -np 16 hostname<br> It gives output of total 16 hostname, i.e run it on 4 tasks on each of 4 nodes<br>2. If in job scripts use sleep like this<br>
</font><font face="verdana,sans-serif"> #PBS -l nodes=4,tpn=4<br>
mpirun -np 16 sleep 200<br> It shows running only 4 task on 4 different nodes.<br><br>So it looks like me it run the tasks only on single slot out of 4 slots.<br><br>Please advise if there anything needs to be change in the config.<br>
<br>Regards<br>Govind<br></font><font face="verdana,sans-serif"><br></font><br><div class="gmail_quote">On Wed, Nov 17, 2010 at 9:50 PM, <span dir="ltr"><<a href="mailto:pat.o%27bryant@exxonmobil.com">pat.o'bryant@exxonmobil.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;"><br>
<br>
Govind,<br>
Go to the Adaptive Resource web page and get the Torque manual in PDF<br>
format. Next search for "tpn" which stands for "task-per-node". There is an<br>
explanation of how "ppn" and "tpn" are different.<br>
>From your test cases it is as though your jobs are being interpreted in a<br>
task fashion and not a node fashion. So, when the statement "nodes=3:ppn=1"<br>
is made, this is a request for (3 x 1) tasks instead of what you intended.<br>
Try this instead: "nodes=3,tpn=1". The request says "3 nodes" with "1 task<br>
per node". Note that there is a "comma" after the "nodes" values and not a<br>
":". Hopefully you will get a better result. The use of ppn can be<br>
confusing.<br>
Thanks,<br>
Pat<br>
<br>
<br>
J.W. (Pat) O'Bryant,Jr.<br>
Business Line Infrastructure<br>
Technical Systems, HPC<br>
<br>
<br>
<br>
<br>
<br>
Govind<br>
<govind.rhul@g<br>
<a href="http://ooglemail.com" target="_blank">ooglemail.com</a>> To<br>
Sent by: Torque Users Mailing List<br>
torqueusers-bo <<a href="mailto:torqueusers@supercluster.org">torqueusers@supercluster.org</a>><br>
unces@superclu cc<br>
<a href="http://ster.org" target="_blank">ster.org</a><br>
Subject<br>
Re: [torqueusers] strange behaviour<br>
11/17/10 09:51 of ppn<br>
AM<br>
<br>
<br>
Please respond<br>
to<br>
Torque Users<br>
Mailing List<br>
<torqueusers@s<br>
upercluster.or<br>
g><br>
<div><div></div><div class="h5"><br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
Hi Brian,<br>
<br>
I do'nt want to block a complete node for a single job.<br>
My requirement is to request multiple processor on different nodes which is<br>
not working at the moment.<br>
<br>
Thanks<br>
Govind<br>
<br>
<br>
On Mon, Nov 15, 2010 at 5:53 PM, Andrus, Brian Contractor <<a href="mailto:bdandrus@nps.edu">bdandrus@nps.edu</a><br>
> wrote:<br>
Govind,<br>
<br>
You may want to add:<br>
<br>
#PBS -l naccesspolicy=singlejob<br>
<br>
This will cause allocation to be a single job per node.<br>
Given your resource request, you do not specify you need exclusive use of<br>
the node, so as far as torque is concerned, there are processors<br>
available to be assigned to jobs, which it does.<br>
<br>
Brian Andrus<br>
<br>
<br>
<br>
________________________________<br>
<br>
From: <a href="mailto:torqueusers-bounces@supercluster.org">torqueusers-bounces@supercluster.org</a> on behalf of Govind Songara<br>
Sent: Fri 11/12/2010 8:25 AM<br>
To: Torque Users Mailing List<br>
Subject: [torqueusers] strange behaviour of ppn<br>
<br>
<br>
Hi,<br>
<br>
<br>
I am not expert on torque configurations, so might something wrong with<br>
configurations.<br>
I am seeing a strange behaviour of ppn variable.<br>
My nodes config is something like<br>
node01 np=4<br>
node02 np=4<br>
<br>
snippet of maui config<br>
JOBNODEMATCHPOLICY EXACTNODE<br>
ENABLEMULTINODEJOBS TRUE<br>
NODEACCESSPOLICY SHARED<br>
<br>
<br>
snippet of queue config resources_available.nodect =<br>
65 resources_assigned.nodect = 5 resources_default.nodes = 1<br>
<br>
sample script<br>
------------------------------------<br>
#PBS -q long<br>
#PBS -l nodes=2:ppn=1<br>
<br>
echo This jobs runs on the following processors:<br>
echo `cat $PBS_NODEFILE`<br>
NPROCS=`wc -l < $PBS_NODEFILE`<br>
echo This job has allocated $NPROCS processors<br>
hostname<br>
------------------------------------<br>
<br>
Below is my result in the tables<br>
<br>
<br>
<br>
nodes<br>
<br>
ppn<br>
<br>
no. process run (hostname)<br>
<br>
no. pf processor allocated<br>
<br>
3<br>
<br>
1<br>
<br>
1<br>
<br>
3<br>
<br>
3<br>
<br>
2<br>
<br>
1<br>
<br>
2<br>
<br>
3<br>
<br>
3<br>
<br>
1<br>
<br>
3<br>
<br>
3<br>
<br>
4<br>
<br>
1<br>
<br>
4<br>
<br>
In case 1, it gives 3 processor on same node which is incorrect, it<br>
should give 1 processor on 3 different nodes<br>
In case2, it give only 2 processor on same node, it should 2 processor on<br>
3 different nodes (total 6 processor) and similar behaviour with the last<br>
tow cases.<br>
In all the cases the hostname command run only once, which should run at<br>
least on total number of allocated processors.<br>
<br>
<br>
Due to this strange behaviour i can not run mpi jobs correctly, kindly<br>
advise on this problem.<br>
<br>
TIA<br>
<br>
Regards<br>
Govind<br>
<br>
<br>
<br>
<br>
<br>
<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>
<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>
<br>
<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>
</div></div></blockquote></div><br>