Hi, > Am 19.05.2017 um 18:10 schrieb Marco Donauer <mdona...@univa.com>: > > Hi Juan, > > sure this is right. We always recommend to try to keep the number of queues > low and if possible problems, which can be resolved with an other solution > shouldn't be resolved via setting up unnecessary queues.
Just for curiosity: was the number of scratch directories in some way enhanced in Univa GE? As I mentioned: I use different queues only to have different locations of the $TMPDIR on one and the same node (depending on the job requirements: traditional HD, SSD or RAM-Disk). Otherwise I could live with one queue. https://arc.liv.ac.uk/trac/SGE/ticket/1290 — Reuti > Each queue and this is not just a problem with Univa Grid Engine increases > the > computational load for the scheduler. In small clusters maybe not a problem, > but running it in huge compute clusters, with many 100k of jobs, then each > queue has a negative impact. For smaller cluster with just a few 10s hosts or > 10k of jobs, not really visible. > An other point is, that a low number of queues, means, less management and > administration effort. > The best case is to have 1 cluster queue. > > Regards, > Marco > > > On Friday 19 May 2017 15:52:36 juanesteban.jime...@mdc-berlin.de wrote: >> Me, nothing. My colleagues attended your training and were told not to >> create individual queues for solving issues like gpu node access. They are >> using Univa on their cluster, I am not. >> >> Mfg, >> Juan Jimenez >> System Administrator, HPC >> MDC Berlin / IT-Dept. >> Tel.: +49 30 9406 2800 >> >> >> ________________________________________ >> From: Marco Donauer [mdona...@univa.com] >> Sent: Friday, May 19, 2017 17:49 >> To: Jimenez, Juan Esteban; Reuti >> Cc: SGE-discuss@liv.ac.uk >> Subject: Re: [SGE-discuss] GPUs as a resource >> >> Hi Juan, >> >> what have you been told by Univa, >> regarding GPUs? >> >> Regards >> Marco >> >> >> Am 19. Mai 2017 17:12:42 schrieb "juanesteban.jime...@mdc-berlin.de" >> >> <juanesteban.jime...@mdc-berlin.de>: >>> I am just telling you what my colleagues say they were told by Univa. >>> >>> Mfg, >>> Juan Jimenez >>> System Administrator, HPC >>> MDC Berlin / IT-Dept. >>> Tel.: +49 30 9406 2800 >>> >>> >>> ________________________________________ >>> From: Reuti [re...@staff.uni-marburg.de] >>> Sent: Friday, May 19, 2017 17:01 >>> To: Jimenez, Juan Esteban >>> Cc: William Hay; SGE-discuss@liv.ac.uk >>> Subject: Re: [SGE-discuss] GPUs as a resource >>> >>>> Am 19.05.2017 um 16:35 schrieb juanesteban.jime...@mdc-berlin.de: >>>>> You are being told by who or what? If it is a what then the exact >>>>> message >>>>> is helpful? >>>> >>>> By my colleagues who are running a 2nd cluster using Univa GridEngine. >>>> This >>>> was a warning from Univa not to do it that way because it increases >>>> qmaster >>>> workload >>> >>> I second queue increases the workload? I don't think that this is >>> noticeable. We have several queues for several $TMPDIR settings and it >>> wasn't a problem up to now. And I also don't see a high load on this >>> machine because of this setting. >>> >>> -- Reuti >>> >>>> Juan >>>> _______________________________________________ >>>> SGE-discuss mailing list >>>> SGE-discuss@liv.ac.uk >>>> https://arc.liv.ac.uk/mailman/listinfo/sge-discuss >>> >>> _______________________________________________ >>> SGE-discuss mailing list >>> SGE-discuss@liv.ac.uk >>> https://arc.liv.ac.uk/mailman/listinfo/sge-discuss > > -- > Marco Donauer | Senior Software Engineer & Support Manager > Univa Corporation Regensburg/Germany > E: mdona...@univa.com| D: +1.512.782.4453 | M: +49.151.466.396.92 > Toll-Free +1.800.370.5320 ext.1015 > W: Univa.com | S: support.univa.com | T: twitter.com/Grid_Engine > _______________________________________________ SGE-discuss mailing list SGE-discuss@liv.ac.uk https://arc.liv.ac.uk/mailman/listinfo/sge-discuss
