<div dir="ltr">Ahh sorry, my sites file was missing slots value. Swift was using the default slots value of 20. Rerunning with slots=1 gives the desired task rate.<div><br></div><div>Thanks,</div><div>Ketan</div></div><div class="gmail_extra">
<br><br><div class="gmail_quote">On Tue, Oct 1, 2013 at 2:42 PM, Mihael Hategan <span dir="ltr"><<a href="mailto:hategan@mcs.anl.gov" target="_blank">hategan@mcs.anl.gov</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
slots = 1?<br>
<div><div class="h5"><br>
On Tue, 2013-10-01 at 14:35 -0500, Ketan Maheshwari wrote:<br>
> Hi,<br>
><br>
> On stampede, I am observing that the sites parameters such as ppn and<br>
> jobspernode are being overridden by the throttle value. I intended to run 1<br>
> job per node on a single node run. However, it seems more jobs are running.<br>
> My sites parameters are as follows:<br>
><br>
> <profile namespace="globus" key="jobsPerNode">1</profile><br>
> <profile namespace="globus" key="ppn">1</profile><br>
> <profile namespace="globus" key="maxTime">7500</profile><br>
> <profile namespace="globus" key="maxwalltime">00:10:00</profile><br>
> <profile namespace="globus" key="lowOverallocation">100</profile><br>
> <profile namespace="globus" key="highOverallocation">100</profile><br>
> <profile namespace="globus" key="queue">normal</profile><br>
> <profile namespace="globus" key="nodeGranularity">1</profile><br>
> <profile namespace="globus" key="maxNodes">1</profile><br>
> <profile namespace="karajan" key="jobThrottle">.3199</profile><br>
><br>
> To reproduce, I run a catsnsleep example with 100 tasks and sleep value 10<br>
> seconds. From the standard out and completion time, it seems Swift is<br>
> running more jobs in parallel despite jobsPerNode and ppn set to be 1.<br>
><br>
> Attached is the stdout and log for this run. I am setting throttle high<br>
> since otherwise coasters will leave resources and try to reacquire on each<br>
> job burst.<br>
><br>
> Thanks,<br>
> Hi,<br>
><br>
><br>
> On stampede, I am observing that the sites parameters such as ppn and<br>
> jobspernode are being overridden by the throttle value. I intended to<br>
> run 1 job per node on a single node run. However, it seems more jobs<br>
> are running. My sites parameters are as follows:<br>
><br>
><br>
> <profile namespace="globus" key="jobsPerNode">1</profile><br>
> <profile namespace="globus" key="ppn">1</profile><br>
> <profile namespace="globus" key="maxTime">7500</profile><br>
> <profile namespace="globus" key="maxwalltime">00:10:00</profile><br>
> <profile namespace="globus" key="lowOverallocation">100</profile><br>
> <profile namespace="globus" key="highOverallocation">100</profile><br>
> <profile namespace="globus" key="queue">normal</profile><br>
> <profile namespace="globus" key="nodeGranularity">1</profile><br>
> <profile namespace="globus" key="maxNodes">1</profile><br>
> <profile namespace="karajan" key="jobThrottle">.3199</profile><br>
><br>
><br>
> To reproduce, I run a catsnsleep example with 100 tasks and sleep<br>
> value 10 seconds. From the standard out and completion time, it seems<br>
> Swift is running more jobs in parallel despite jobsPerNode and ppn set<br>
> to be 1.<br>
><br>
><br>
> Attached is the stdout and log for this run. I am setting throttle<br>
> high since otherwise coasters will leave resources and try to<br>
> reacquire on each job burst.<br>
><br>
><br>
> Thanks,<br>
> --<br>
> Ketan<br>
><br>
><br>
</div></div>> _______________________________________________<br>
> Swift-devel mailing list<br>
> <a href="mailto:Swift-devel@ci.uchicago.edu">Swift-devel@ci.uchicago.edu</a><br>
> <a href="https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-devel" target="_blank">https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-devel</a><br>
<br>
<br>
</blockquote></div><br><br clear="all"><div><br></div>-- <br><font face="'courier new', monospace">Ketan</font><br><br>
</div>