Sorry, I forgot to forward this solution to the list... <br><br><div class="gmail_quote">---------- Forwarded message ----------<br>From: <b class="gmail_sendername">carlos vasco</b> <<a href="mailto:carles.vasco@gmail.com">
carles.vasco@gmail.com</a>><br>Date: Dec 13, 2007 3:28 PM<br>Subject: Re: [torqueusers] Submiting jobs to multicore/processor nodes<br>To: Glen Beane <<a href="mailto:glen.beane@gmail.com">glen.beane@gmail.com</a>>
<br><br><br>Great, it works now.<br><br>I have specified<br><pre>ENABLEMULTIREQJOBS TRUE<br><br>on the maui.cfg configuration file.<br><br>Thanks a lot,<br><font color="#888888">Carlos<br></font></pre><div><div></div>
<div class="Wj3C7c"><br><br><div class="gmail_quote">On Dec 13, 2007 3:18 PM, Glen Beane <
<a href="mailto:glen.beane@gmail.com" target="_blank">glen.beane@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
That is a scheduler configuration issue, not a torque issue.
<br><br>your scheduler has put a hold on the job because it is configured to not allow multi-request PBS jobs. I believe in Maui you need to enable this - It should work fine "out of the box" with Moab or the simple FIFO torque scheduler.
<div><div></div><div><br><br><div class="gmail_quote">On Dec 13, 2007 9:16 AM, carlos vasco <<a href="mailto:carles.vasco@gmail.com" target="_blank">carles.vasco@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>It doesn't work either, I tried it. checkjob tells that:<br><br>IWD: [NONE] Executable: [NONE]<br>Bypass: 0 StartCount: 0<br>PartitionMask: [ALL]<br>Holds: Batch (hold reason: PolicyViolation)<br>Messages: multi-req PBS jobs not allowed
<br>PE: 6.00 StartPriority: 1<br>cannot select job 6245 for partition DEFAULT (job hold active)<br><br>Regards,<br><font color="#888888">Carlos</font><div><div></div><div><br><br><br><div class="gmail_quote">
On Dec 13, 2007 3:08 PM, Glen Beane <<a href="mailto:glen.beane@gmail.com" target="_blank">
glen.beane@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div><br><br><div class="gmail_quote">
On Dec 13, 2007 5:15 AM, carlos vasco <<a href="mailto:carles.vasco@gmail.com" target="_blank">carles.vasco@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Dear all,<br><br>We have a cluster with SMP/dual core nodes, each one has two processors with two cores each. In the nodes file we have 4 virtual processors defined.<br><br>We use to submit 2,4 and 6 processes jobs, and we find more efficient to pack the jobs into a node, because of our slow network.
<br>The 2 and 4 cases are easy and we put in our scritpt:<br><br>#PBS -l nodes=1:ppn=2 <br>#PBS -l nodes=1:ppn=4<br><br>but the case of 6, we would like to run 4 in one node and 2 in other node, doesn't run. We, following the man page, put:
<br>#PBS -l nodes=1:ppn=4+ppn=2<br><br>Any idea of how to impose that configuration?<br></blockquote></div><br></div>#PBS -l nodes=1:ppn=4+1:ppn=2<br>
</blockquote></div><br>
</div></div></blockquote></div><br>
</div></div></blockquote></div><br>
</div></div></div><br>