On Tue, Feb 17, 2015 at 11:33 AM, Sean Last <qkth...@gmail.com> wrote:
> Is there a good metric to reduce the number of executors by? > > My heuristic is that I try to keep the number of executors down to no more than one less than the number of processors on the computer. However, another heuristic might be to watch the memory use of your jobs and limit the number of executors so that your jobs never exceed the physical memory available on the system. I'm sure others have different heuristics based on their job types. My job types tend to be more CPU bound than memory bound. Mark Waite > On Tuesday, February 17, 2015 at 1:18:29 PM UTC-5, Mark Waite wrote: >> >> I think that message usually means you've exhausted virtual memory. With >> most systems, if you've exhausted virtual memory, you've already been >> suffering poor performance due to thrashing of your paging system (swap >> space). >> >> Reduce the number of executors, or increase the physical memory and >> possibly the size of the paging file(s). >> >> Mark Waite >> >> On Tue, Feb 17, 2015 at 10:45 AM, Sean Last <qkt...@gmail.com> wrote: >> >>> Thu Jan 29 15:48:45 2015] Out of memory: Kill process 19422 (java) >>> score 139 or sacrifice child >>> [Thu Jan 29 18:24:25 2015] Out of memory: Kill process 19422 (java) >>> score 140 or sacrifice child >>> [Thu Jan 29 20:23:21 2015] Out of memory: Kill process 19422 (java) >>> score 143 or sacrifice child >>> [Fri Jan 30 20:25:04 2015] Out of memory: Kill process 19422 (java) >>> score 162 or sacrifice child >>> [Tue Feb 3 19:19:10 2015] Out of memory: Kill process 19422 (java) >>> score 163 or sacrifice child >>> [Wed Feb 4 14:16:41 2015] Out of memory: Kill process 19422 (java) >>> score 165 or sacrifice child >>> [Wed Feb 4 16:22:11 2015] Out of memory: Kill process 19422 (java) >>> score 166 or sacrifice child >>> [Wed Feb 4 19:20:27 2015] Out of memory: Kill process 19422 (java) >>> score 163 or sacrifice child >>> [Fri Feb 6 14:16:21 2015] Out of memory: Kill process 25783 (java) >>> score 112 or sacrifice child >>> [Mon Feb 9 06:20:54 2015] Out of memory: Kill process 31175 (java) >>> score 140 or sacrifice child >>> [Mon Feb 9 08:55:12 2015] Out of memory: Kill process 25783 (java) >>> score 115 or sacrifice child >>> [Mon Feb 9 16:55:25 2015] Out of memory: Kill process 25783 (java) >>> score 121 or sacrifice child >>> [Mon Feb 9 16:55:59 2015] Out of memory: Kill process 25783 (java) >>> score 121 or sacrifice child >>> [Mon Feb 9 17:01:16 2015] Out of memory: Kill process 25783 (java) >>> score 121 or sacrifice child >>> [Wed Feb 11 10:13:51 2015] Out of memory: Kill process 25783 (java) >>> score 127 or sacrifice child >>> [Wed Feb 11 10:16:37 2015] Out of memory: Kill process 25783 (java) >>> score 127 or sacrifice child >>> >>> Should I reduce the number of executors, or how can I solve this? >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "Jenkins Users" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to jenkinsci-use...@googlegroups.com. >>> To view this discussion on the web visit https://groups.google.com/d/ >>> msgid/jenkinsci-users/4915c03a-d298-4059-b410- >>> d14e2a9d981e%40googlegroups.com >>> <https://groups.google.com/d/msgid/jenkinsci-users/4915c03a-d298-4059-b410-d14e2a9d981e%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> >> >> -- >> Thanks! >> Mark Waite >> > -- > You received this message because you are subscribed to the Google Groups > "Jenkins Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to jenkinsci-users+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/jenkinsci-users/9b95be75-2f69-43d8-8fdc-af0e04a7baf3%40googlegroups.com > <https://groups.google.com/d/msgid/jenkinsci-users/9b95be75-2f69-43d8-8fdc-af0e04a7baf3%40googlegroups.com?utm_medium=email&utm_source=footer> > . > For more options, visit https://groups.google.com/d/optout. > -- Thanks! Mark Waite -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-users/CAO49JtHr0mgNue0vLU8n-qyC5ZdV%2BmOdX0OLVtpzC75%2B%3Dh_5Yg%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
