David,<div><br></div><div><div class="gmail_quote">On Thu, Oct 20, 2011 at 10:37 AM, David Kelly <span dir="ltr"><<a href="mailto:davidk@ci.uchicago.edu">davidk@ci.uchicago.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Ketan,<br>
<br>
I think you're right - I believe that line should be:<br>
<br>
#$ -pe <parallel_environment> <maxNodes*nodeGranularity><br></blockquote><div><br></div><div>I am not sure, it should be maxnodes*nodegranularity, since nodegranularity means the nodes-to-be-packed per job. I think, this should be,</div>
<div>maxnodes*corespernode, which could be a static constant (value 16) for ranger. I could be wrong here, not sure.</div><div><br></div><div>Furthermore, it seems the parallel_environment tag of sites.xml is still not being honored. It always defaults to '1way', irrespective of the provided value. I am using 0.93RC3.</div>
<div><br></div><div>May be we can debug this together.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<br>
I'll add this and do a little more testing to see if I can reproduce Sarah's problem.<br>
<font color="#888888"><br>
David<br>
</font><div><div></div><div class="h5"><br>
> I have a little confusion here: the desired line in the final pbs<br>
> script should be : #$ -pe <n>way 256; in order to have 256 procs,<br>
> however, putting maxnodes=16 on sites.xml results in the following<br>
> line on pbs:<br>
> #$ -pe <n>way 16;<br>
> I understand this number 16/256 is for procs since, when putting 256<br>
> with development queue, ranger indeed allows the job to run in<br>
> development queue.<br>
><br>
><br>
><br>
> I would also suggest (unless you have already done this) that you test<br>
> first on a very small run (like a single RInvoke app call) and then<br>
> scale up to just a few voxels per dataset before trying such a large<br>
> run. Have you already tested that?<br>
><br>
> Lastly, when reporting problems like this, the swift standard<br>
> output/err is also very helpful to get a higher-level view of what<br>
> went wrong.<br>
><br>
> Swift needs to clearly return errors from the local resource provider,<br>
> which it doesnt seem to be doing here. Ive filed this as bug 593 and<br>
> assigned to David.<br>
><br>
> Please let us know if changing the queue and/or slots resolves the<br>
> problem. As mentioned in the bug report I think you can set debug=true<br>
> (or yes?) in the provider-sge.properties file and get swift to<br>
> preserve the output from SGE in ~/.globus/scripts. (In fact that may<br>
> already be preserved, I am not sure). Please check there to see if the<br>
> SGE error is there.<br>
><br>
> Thanks,<br>
><br>
> - Mike<br>
><br>
><br>
><br>
> ----- Original Message -----<br>
> > From: "Sarah Kenny" < <a href="mailto:skenny@uchicago.edu">skenny@uchicago.edu</a> ><br>
><br>
> > To: "Mihael Hategan" < <a href="mailto:hategan@mcs.anl.gov">hategan@mcs.anl.gov</a> ><br>
> > Cc: "Anjali Raja" < <a href="mailto:anjraja@gmail.com">anjraja@gmail.com</a> >, "Swift Devel" <<br>
> > <a href="mailto:swift-devel@ci.uchicago.edu">swift-devel@ci.uchicago.edu</a> >, "Swift User"<br>
> > < <a href="mailto:swift-user@ci.uchicago.edu">swift-user@ci.uchicago.edu</a> ><br>
> > Sent: Thursday, October 20, 2011 6:07:09 AM<br>
> > Subject: Re: [Swift-devel] [Swift-user] gram on ranger<br>
><br>
> > hi all, one of our users, anjali (cc'd here) is trying to submit<br>
> > this<br>
> > ~400k job workflow to ranger...thought i'd see if you felt like<br>
> > having<br>
> > a look :)<br>
> ><br>
> > log is here:<br>
> > /home/skenny/swift_logs/corr_multisubj-20111018-1321-ihf8hz5g.log<br>
> ><br>
> > sites file:<br>
> ><br>
> > <config><br>
> > <pool handle="RANGER"><br>
> > <execution provider="coaster" jobManager="gt2:SGE" url="<br>
> > <a href="http://gatekeeper.ranger.tacc.teragrid.org" target="_blank">gatekeeper.ranger.tacc.teragrid.org</a> "/><br>
><br>
><br>
><br>
> > <filesystem provider="gsiftp" url="gsiftp://<br>
> > <a href="http://gridftp.ranger.tacc.teragrid.org" target="_blank">gridftp.ranger.tacc.teragrid.org</a> "/><br>
> > <profile namespace="globus" key="maxtime">7200</profile><br>
> > <profile namespace="globus" key="maxWallTime">00:20:00</profile><br>
> > <profile namespace="globus" key="jobsPerNode">1</profile><br>
> > <profile namespace="globus" key="nodeGranularity">64</profile><br>
> > <profile namespace="globus" key="maxNodes">256</profile><br>
> > <profile namespace="globus" key="queue">development</profile><br>
> > <profile namespace="karajan" key="jobThrottle">1.28</profile><br>
> > <profile namespace="globus" key="project">TG-DBS080004N</profile><br>
> > <profile namespace="globus" key="pe">16way</profile><br>
> > <profile namespace="karajan" key="initialScore">10000</profile><br>
> > <workdirectory>/work/00926/tg459516/swiftwork</workdirectory><br>
> > </pool><br>
> > </config><br>
> ><br>
> ><br>
> > On Wed, Oct 12, 2011 at 12:13 PM, Mihael Hategan <<br>
> > <a href="mailto:hategan@mcs.anl.gov">hategan@mcs.anl.gov</a><br>
> > > wrote:<br>
> ><br>
> ><br>
> ><br>
> > On Tue, 2011-10-11 at 17:13 -0700, Sarah Kenny wrote:<br>
> > ><br>
> > ><br>
> > > On Tue, Oct 11, 2011 at 4:23 PM, Mihael Hategan <<br>
> > > <a href="mailto:hategan@mcs.anl.gov">hategan@mcs.anl.gov</a> ><br>
> > > wrote:<br>
> > > Is this with a persistent coaster service?<br>
> > ><br>
> > > admittedly i have not used persistent coaster service...should i?<br>
> ><br>
> > No. I was just trying to figure out whether it might be something<br>
> > related to the persistent version.<br>
> ><br>
> ><br>
> ><br>
> ><br>
> > > i feel like it's documented *somewhere* (?)<br>
> > ><br>
> > > for now i've tried setting 'sitedir.keep=true' in the config so<br>
> > > maybe<br>
> > > it won't try to run the cleanup job...we'll see (waiting in q)<br>
> > ><br>
> > ><br>
> > ><br>
> > > On Tue, 2011-10-11 at 12:05 -0700, Sarah Kenny wrote:<br>
> > > ><br>
> > > ><br>
> > > > On Tue, Oct 11, 2011 at 11:49 AM, David Kelly<br>
> > > < <a href="mailto:davidk@ci.uchicago.edu">davidk@ci.uchicago.edu</a> ><br>
> > > > wrote:<br>
> > > ><br>
> > > > That could be it.. maybe a cleanup script is not<br>
> > > getting the<br>
> > > > right parameters and failing. Do you happen to have<br>
> > > a copy of<br>
> > > > the coaster log?<br>
> > > ><br>
> > > > just put it in /home/skenny/swift_logs<br>
> > > ><br>
> > > ><br>
> > > > Maybe there will be some clues in there.<br>
> > > ><br>
> > > > ----- Original Message -----<br>
> > > > > From: "Sarah Kenny" < <a href="mailto:skenny@uchicago.edu">skenny@uchicago.edu</a> ><br>
> > > ><br>
> > > > > To: "David Kelly" < <a href="mailto:davidk@ci.uchicago.edu">davidk@ci.uchicago.edu</a> ><br>
> > > > > Cc: "Swift Devel" < <a href="mailto:swift-devel@ci.uchicago.edu">swift-devel@ci.uchicago.edu</a> >,<br>
> > > "Swift<br>
> > > > User" < <a href="mailto:swift-user@ci.uchicago.edu">swift-user@ci.uchicago.edu</a> >, "Justin M<br>
> > > Wozniak"<br>
> > > > > < <a href="mailto:wozniak@mcs.anl.gov">wozniak@mcs.anl.gov</a> ><br>
> > > ><br>
> > > > > Sent: Tuesday, October 11, 2011 1:32:37 PM<br>
> > > > > Subject: Re: [Swift-user] gram on ranger<br>
> > > ><br>
> > > > > so, this workflow completes all the jobs but then<br>
> > > just hangs<br>
> > > > > indefinitely at the end...maybe a stray cleanup<br>
> > > job?<br>
> > > > ><br>
> > > > > log is here:<br>
> > > > ><br>
> > > ><br>
> > > > /home/skenny/swift_logs/corr-20111010-2104-fl5yngd9.log<br>
> > > > ><br>
> > > > > just tweaked the sites file a bit from what david<br>
> > > sent me:<br>
> > > > ><br>
> > > > > <config><br>
> > > > > <pool handle="RANGER"><br>
> > > > > <execution provider="coaster" jobManager="gt2:SGE"<br>
> > > url="<br>
> > > > > <a href="http://gatekeeper.ranger.tacc.teragrid.org" target="_blank">gatekeeper.ranger.tacc.teragrid.org</a> "/><br>
> > > > > <filesystem provider="gsiftp" url="gsiftp://<br>
> > > ><br>
> > > > > <a href="http://gridftp.ranger.tacc.teragrid.org" target="_blank">gridftp.ranger.tacc.teragrid.org</a> "/><br>
> > > ><br>
> > > > > <profile namespace="globus"<br>
> > > key="maxtime">28800</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="maxWallTime">00:15:00</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="jobsPerNode">1</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="nodeGranularity">64</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="maxNodes">256</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="queue">normal</profile><br>
> > > > > <profile namespace="karajan"<br>
> > > key="jobThrottle">1</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="project">TG-DBS080004N</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="pe">16way</profile><br>
> > > > > <profile namespace="karajan"<br>
> > > > key="initialScore">10000</profile><br>
> > > > ><br>
> > > ><br>
> > > <workdirectory>/work/00043/tg457040/sidgrid_out/skenny</workdirectory><br>
> > > > > </pool><br>
> > > > > </config><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > On Mon, Oct 10, 2011 at 3:43 PM, Sarah Kenny <<br>
> > > > <a href="mailto:skenny@uchicago.edu">skenny@uchicago.edu</a> ><br>
> > > > > wrote:<br>
> > > > ><br>
> > > > ><br>
> > > > > ok, thanks, got in the queue now...also, realized<br>
> > > my last<br>
> > > > run may have<br>
> > > > > been using the old swift. apparently i had<br>
> > > SWIFT_HOME set in<br>
> > > > my env<br>
> > > > > and that overrides the newer swift i had set in my<br>
> > > PATH.<br>
> > > > ><br>
> > > > > ~sk<br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > On Mon, Oct 10, 2011 at 12:28 PM, David Kelly <<br>
> > > > <a href="mailto:davidk@ci.uchicago.edu">davidk@ci.uchicago.edu</a><br>
> > > > > > wrote:<br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > Sarah,<br>
> > > > ><br>
> > > > > Can you give this another try with the latest<br>
> > > 0.93? I made<br>
> > > > some<br>
> > > > > changes to the coaster and sge providers and was<br>
> > > able to get<br>
> > > > it<br>
> > > > > working with a simple catns script. Here is the<br>
> > > > configuration file I<br>
> > > > > was using:<br>
> > > > ><br>
> > > > > <config><br>
> > > > > <pool handle="ranger"><br>
> > > > > <execution provider="coaster" jobManager="gt2:SGE"<br>
> > > url="<br>
> > > > > <a href="http://gatekeeper.ranger.tacc.teragrid.org" target="_blank">gatekeeper.ranger.tacc.teragrid.org</a> "/><br>
> > > > ><br>
> > > > > <filesystem provider="gsiftp" url="gsiftp://<br>
> > > ><br>
> > > > > <a href="http://gridftp.ranger.tacc.teragrid.org" target="_blank">gridftp.ranger.tacc.teragrid.org</a> "/><br>
> > > ><br>
> > > > > <profile namespace="globus"<br>
> > > key="maxtime">3600</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="maxWallTime">00:00:03</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="jobsPerNode">1</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="nodeGranularity">16</profile><br>
> > > > > <profile namespace="globus"<br>
> > > key="maxNodes">16</profile><br>
> > > > > <profile namespace="globus"<br>
> > > > key="queue">development</profile><br>
> > > > > <profile namespace="karajan"<br>
> > > key="jobThrottle">0.9</profile><br>
> > > > ><br>
> > > > > <profile namespace="globus"<br>
> > > > key="project">TG-DBS080004N</profile><br>
> > > > ><br>
> > > > > <profile namespace="globus"<br>
> > > key="pe">16way</profile><br>
> > > > ><br>
> > > ><br>
> > > <workdirectory>/share/home/01503/davidkel/swiftwork</workdirectory><br>
> > > > > </pool><br>
> > > > > </config><br>
> > > > ><br>
> > > > > Thanks,<br>
> > > > ><br>
> > > > > David<br>
> > > > ><br>
> > > > > ----- Original Message -----<br>
> > > > ><br>
> > > > > > From: "Sarah Kenny" < <a href="mailto:skenny@uchicago.edu">skenny@uchicago.edu</a> ><br>
> > > > > > To: "Justin M Wozniak" < <a href="mailto:wozniak@mcs.anl.gov">wozniak@mcs.anl.gov</a> ><br>
> > > > > > Cc: "Swift Devel" < <a href="mailto:swift-devel@ci.uchicago.edu">swift-devel@ci.uchicago.edu</a><br>
> > > >, "Swift<br>
> > > > User" <<br>
> > > > > > <a href="mailto:swift-user@ci.uchicago.edu">swift-user@ci.uchicago.edu</a> ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > > Sent: Friday, October 7, 2011 3:13:57 PM<br>
> > > > > > Subject: Re: [Swift-user] gram on ranger<br>
> > > > ><br>
> > > > /home/skenny/swift_logs/dummy-20111005-0126-6575n7x5.log<br>
> > > > > ><br>
> > > > > > on ci<br>
> > > > > ><br>
> > > > > ><br>
> > > > > > On Fri, Oct 7, 2011 at 8:16 AM, Justin M Wozniak<br>
> > > <<br>
> > > > > > <a href="mailto:wozniak@mcs.anl.gov">wozniak@mcs.anl.gov</a><br>
> > > > > > > wrote:<br>
> > > > > ><br>
> > > > > ><br>
> > > > > ><br>
> > > > > > Can I take a look at the log?<br>
> > > > > ><br>
> > > > > ><br>
> > > > > ><br>
> > > > > ><br>
> > > > > > On Thu, 6 Oct 2011, Sarah Kenny wrote:<br>
> > > > > ><br>
> > > > > ><br>
> > > > > ><br>
> > > > > > hey all, i'm trying to submit to gram on ranger<br>
> > > using the<br>
> > > > latest<br>
> > > > > > swift<br>
> > > > > > (built from trunk). it failes like so:<br>
> > > > > ><br>
> > > > > > Cannot submit job<br>
> > > > > > Caused by:<br>
> > > > > > org.globus.cog.abstraction. impl.common.task.<br>
> > > > > > TaskSubmissionException:<br>
> > > > > > Cannot<br>
> > > > > > submit job<br>
> > > > > > Caused by: org.globus.gram.GramException:<br>
> > > Parameter not<br>
> > > > supported<br>
> > > > > > Cannot submit job<br>
> > > > > ><br>
> > > > > > the gram log was saying first that 'jobsPerNode'<br>
> > > is not<br>
> > > > supported so<br>
> > > > > > i<br>
> > > > > > changed it to workersPerNode and then it was<br>
> > > saying<br>
> > > > 'maxnodes' is<br>
> > > > > > not<br>
> > > > > > supported. here's my sites file:<br>
> > > > > ><br>
> > > > > > <config><br>
> > > > > > <pool handle="RANGER"><br>
> > > > > > <profile namespace="karajan"<br>
> > > key="initialScore">10000</<br>
> > > > profile><br>
> > > > > > <profile namespace="karajan"<br>
> > > key="jobThrottle">1</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="maxWallTime">00:15:00</<br>
> > > > profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="maxTime">86400</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="slots">1</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="maxNodes">256</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="pe">16way</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="workersPerNode">1</<br>
> > > > profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="nodeGranularity">64</<br>
> > > > profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="queue">normal</profile><br>
> > > > > > <profile namespace="globus"<br>
> > > key="project">TG-DBS080004N</<br>
> > > > profile><br>
> > > > > > <filesystem provider="gsiftp" url="gsiftp://<br>
> > > > > > gridftp.ranger.tacc.teragrid. org "/><br>
> > > > ><br>
> > > > > > <execution provider="coaster"<br>
> > > jobManager="gt2:gt2:SGE"<br>
> > > > url="<br>
> > > > > > gatekeeper.ranger.tacc. <a href="http://teragrid.org" target="_blank">teragrid.org</a> "/><br>
> > > > ><br>
> > > > > > <execution provider="gt2" jobManager="SGE" url="<br>
> > > > > > gatekeeper.ranger.tacc. <a href="http://teragrid.org" target="_blank">teragrid.org</a> "/><br>
> > > > > > <workdirectory>/work/00043/<br>
> > > tg457040</workdirectory><br>
> > > > ><br>
> > > > > > </pool><br>
> > > > > > </config><br>
> > > > > ><br>
> > > > > > thoughts? ideas?<br>
> > > > > ><br>
> > > > > > --<br>
> > > > > > Justin M Wozniak<br>
> > > > > ><br>
> > > > > ><br>
> > > > > ><br>
> > > > > > --<br>
> > > > > > Sarah Kenny<br>
> > > > > > Programmer ~ Brain Circuits Laboratory ~ Rm 2224<br>
> > > Bio Sci<br>
> > > > III<br>
> > > > > > University of California Irvine, Dept. of<br>
> > > Neurology ~<br>
> > > > 773-818-8300<br>
> > > > > ><br>
> > > > > ><br>
> > > > > > _______________________________________________<br>
> > > > > > Swift-user mailing list<br>
> > > > > > <a href="mailto:Swift-user@ci.uchicago.edu">Swift-user@ci.uchicago.edu</a><br>
> > > > > ><br>
> > > ><br>
> > > <a href="https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-user" target="_blank">https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-user</a><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > --<br>
> > > > > Sarah Kenny<br>
> > > > > Programmer ~ Brain Circuits Laboratory ~ Rm 2224<br>
> > > Bio Sci III<br>
> > > > > University of California Irvine, Dept. of<br>
> > > Neurology ~<br>
> > > > 773-818-8300<br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > ><br>
> > > > > --<br>
> > > > > Sarah Kenny<br>
> > > > > Programmer ~ Brain Circuits Laboratory ~ Rm 2224<br>
> > > Bio Sci III<br>
> > > > > University of California Irvine, Dept. of<br>
> > > Neurology ~<br>
> > > > 773-818-8300<br>
> > > ><br>
> > > ><br>
> > > ><br>
> > > ><br>
> > > > --<br>
> > > > Sarah Kenny<br>
> > > > Programmer ~ Brain Circuits Laboratory ~ Rm 2224 Bio Sci III<br>
> > > > University of California Irvine, Dept. of Neurology ~<br>
> > > 773-818-8300<br>
> > > ><br>
> > > > _______________________________________________<br>
> > > > Swift-user mailing list<br>
> > > > <a href="mailto:Swift-user@ci.uchicago.edu">Swift-user@ci.uchicago.edu</a><br>
> > > ><br>
> > > <a href="https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-user" target="_blank">https://lists.ci.uchicago.edu/cgi-bin/mailman/listinfo/swift-user</a><br>
> > ><br>
> > ><br>
> > ><br>
> > ><br>
> > ><br>
> > ><br>
> > > --<br>
> > > Sarah Kenny<br>
> > > Programmer ~ Brain Circuits Laboratory ~ Rm 2224 Bio Sci III<br>
> > > University of California Irvine, Dept. of Neurology ~ 773-818-8300<br>
> > ><br>
> ><br>
> ><br>
> ><br>
> ><br>
> ><br>
> > --<br>
> > Sarah Kenny<br>
> > Programmer ~ Brain Circuits Laboratory ~ Rm 2224 Bio Sci III<br>
> > University of California Irvine, Dept. of Neurology ~ 773-818-8300<br>
> ><br>
> ><br>
><br>
> > _______________________________________________<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>
> Michael Wilde<br>
> Computation Institute, University of Chicago<br>
> Mathematics and Computer Science Division<br>
> Argonne National Laboratory<br>
><br>
><br>
><br>
><br>
> _______________________________________________<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>
><br>
><br>
> --<br>
> Ketan<br>
><br>
><br>
><br>
> _______________________________________________<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>
</div></div></blockquote></div><br><br clear="all"><div><br></div>-- <br>Ketan<br><br><br>
</div>