 |
 |
Hi Gidon, thanks for using my site as an example. However, this looks very fishy to me, since we have CMS and ATLAS jobs of approximately the same duration, so one aborted out of 37 cannot decrease use% to 27... Best regards, Antun ----- E-mail: [log in to unmask] Web: http://scl.phy.bg.ac.yu/ Phone: +381 11 3160260, Ext. 152 Fax: +381 11 3162190 Scientific Computing Laboratory Institute of Physics, Belgrade Serbia and Montenegro ----- ---------- Original Message ----------- From: Gidon Moont <[log in to unmask]> To: [log in to unmask] Sent: Tue, 15 Nov 2005 14:36:58 +0000 Subject: Re: [LCG-ROLLOUT] job monitors > Thanks for pointing out the mistake - a silly error in my scripts. > Now fixed. > > However, I didn't notice because such behaviour is not so odd if you > look even now. Take ce.phy.bg.ac.yu - 36 successful jobs, one > aborted after running - only 27% of computing time was for those 36 successes. > > Cheers > > Gidon > > On Tue, 15 Nov 2005, Jeff Templon wrote: > > > Hi *, > > > > Took a look at the monitor page > > > > http://gridportal.hep.ph.ic.ac.uk/rtm/sites_report.html > > > > today (for NIKHEF of course) and saw something I must not understand. > > For NIKHEF (at 14.51 dutch time) the page said > > > > 519 jobs submitted > > 512 jobs successful > > 2 jobs aborted > > 5 jobs cancelled > > > > looks pretty good, so why does it say that the percent of CPU time doing > > useful work is only 48%? In fact it looks like there must be some logic > > error in the calculation, there are many sites with 100%, then NO sites with > > useful time percentages between 50 and 99 percent, then the rest are spread > > between zero and 49%, with some evidence of skewing towards values of 48 to > > 49%. > > > > For example take ce2.egee.cesga.es ... with 369 submitted jobs, 367 of which > > were successful, still 49% "useful time". > > > > Explanation? > > > > JT > > > > -------------------------------------------------- > > Dr. Gidon Moont > High Energy Physics Group, The Blackett Laboratory > Imperial College London, South Kensington Campus > Prince Consort Road, LONDON SW7 2BW > Tel: +44 (0)207 594 7810 ------- End of Original Message -------