Minh Nguyen wrote: > Hi David, > > On Thu, Aug 13, 2009 at 10:52 PM, Dr. David > Kirkby<david.kir...@onetel.net> wrote: >> Running 'make test' I'm seeing a load average of 13 on my dual processor >> Sun, which it is not exactly happy about! >> >> >> Here's the output from prstat >> >> PID USERNAME SIZE RSS STATE PRI NICE TIME CPU PROCESS/NLWP >> 16735 drkirkby 30M 25M run 45 0 0:37:43 11% maxima/1 >> 28928 drkirkby 34M 29M run 35 0 0:27:01 8.4% maxima/1 >> 8183 drkirkby 105M 62M run 45 0 0:00:03 4.8% python/1 >> 1 root 2896K 1760K run 49 0 0:05:22 1.9% init/1 >> 7724 drkirkby 216M 147M sleep 59 0 0:17:48 1.6% firefox-bin/8 >> 4746 drkirkby 147M 93M sleep 54 0 0:03:19 1.0% python/1 >> 22470 drkirkby 167M 122M run 58 0 0:00:49 0.8% python/1 >> 691 drkirkby 203M 80M sleep 59 0 0:54:05 0.6% Xsun/1 >> 913 drkirkby 140M 42M sleep 59 0 0:03:37 0.5% gnome-terminal/2 >> 8100 drkirkby 8144K 5864K sleep 59 0 0:00:00 0.4% python/1 >> 11708 drkirkby 3504K 2776K cpu0 39 0 0:00:00 0.1% prstat/1 >> 1850 drkirkby 20M 13M sleep 49 0 0:04:24 0.0% sunpcbinary/3 >> 777 noaccess 210M 116M sleep 59 0 0:02:08 0.0% java/18 >> 911 drkirkby 68M 5680K sleep 49 0 0:01:51 0.0% >> gnome-netstatus/1 >> 945 drkirkby 236M 75M sleep 49 0 0:03:07 0.0% java/18 >> Total: 129 processes, 326 lwps, load averages: 12.59, 12.73, 12.24 >> >> Has this been seen before? I noticed an earlier sage used up tons of CPU >> time. No particularly process seems to be using a lot, but I think >> processes are starting and stopping quickly. > > Sometimes when I do "make test" and then press control-C before it > finishes, I would get runaway processes that just keeps on computing > for hours. This requires a manual kill. >
I think I have seen this before. I know I have seen Maxima eat up tons of CPU time. But I have some recollection of an unknown consumer of lots of CPU power. I have a 'dtrace' script somewhere for debugging this sort of behavior. I will try that when it happens next. This got a load average of 25 at one point, and then I decided to kill it. dave The load average climbed to 25 and whilst the machine was still just about usable, I terminated the test. --~--~---------~--~----~------------~-------~--~----~ To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
