Are you using containers ? The default cache configuration looks at /prod/meminfo not /sys/fs/cgroup/.* If the host memory isn't appropriate for you want to use for cache, or if the host memory has changed since installation of ML, you will need to manually configure cache in group settings.
We recommend swap be configured - OOM should not take down a node...it's generally better that it become slow due to swap than go away completely. Particularly as the memory condition may be short lived. *actually seeing the memory allocated inside a container is a known PITA. The move to systemd should help and we are looking at this but explicit config is required for now. Aaron Rosenbaum Sent from my iPhone > On Nov 12, 2016, at 12:00 PM, "[email protected]" > <[email protected]> wrote: > > Send General mailing list submissions to > [email protected] > > To subscribe or unsubscribe via the World Wide Web, visit > http://developer.marklogic.com/mailman/listinfo/general > or, via email, send a message with subject or body 'help' to > [email protected] > > You can reach the person managing the list at > [email protected] > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of General digest..." > > > Today's Topics: > > 1. Re: A question about limiting Marklogic memory and CPU usage. > (Yair Etziony) (Josh Bench) > 2. Re: A question about limiting Marklogic memory and CPU usage. > (Danny Sokolsky) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Fri, 11 Nov 2016 13:54:59 -0700 > From: Josh Bench <[email protected]> > Subject: Re: [MarkLogic Dev General] A question about limiting > Marklogic memory and CPU usage. (Yair Etziony) > To: [email protected] > Message-ID: > <CAAYDkZ8ZOhiU6N6+fHciPudX6U=12ne=retuqnuzajm+wjh...@mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Hi Yair, > > It sounds like you have MarkLogic configured to use too much memory. You > could configure the OOM killer, but first, I would fiddle with the > MarkLogic memory settings described here. You can set limits on your memory > usage per db. > > https://docs.marklogic.com/guide/admin/databases#id_59939 > > On Fri, Nov 11, 2016 at 1:00 PM, <[email protected]> > wrote: > >> Send General mailing list submissions to >> [email protected] >> >> To subscribe or unsubscribe via the World Wide Web, visit >> http://developer.marklogic.com/mailman/listinfo/general >> or, via email, send a message with subject or body 'help' to >> [email protected] >> >> You can reach the person managing the list at >> [email protected] >> >> When replying, please edit your Subject line so it is more specific >> than "Re: Contents of General digest..." >> >> >> Today's Topics: >> >> 1. A question about limiting Marklogic memory and CPU usage. >> (Yair Etziony) >> 2. (no subject) (Yair Etziony) >> >> >> ---------------------------------------------------------------------- >> >> Message: 1 >> Date: Fri, 11 Nov 2016 14:32:20 +0100 >> From: Yair Etziony <[email protected]> >> Subject: [MarkLogic Dev General] A question about limiting Marklogic >> memory and CPU usage. >> To: [email protected] >> Message-ID: >> <CALFBOvFrVVPYeDF3pZySRA8S5wbUKnEruBEqZCveZK_XJB6LxQ@mail. >> gmail.com> >> Content-Type: text/plain; charset="utf-8" >> >> Hi All >> i have a question about ML, this more of DBA or admin question, but i will >> take my shot here. >> >> We run Marklogic on Linux server (Centos 7), there are times when we run >> big processes and the ML daemon would take all memory, in situations like >> that the Linux kernel will kill the process and this is not something we >> can accept in our prod stack. >> the question is what would be the best practice to limit the memory and >> cpu consumption for ML daemon? ( for example using linux cgroups) >> >> Thanks in advance for any answer >> >> bests >> Y >> >> -- >> >> --- >> >> Yair Etziony >> >> LambdaWerk GmbH >> >> Oranienburger Stra?e 87/89 >> >> 10178 Berlin >> >> Phone +49 151 75938044 >> -------------- next part -------------- >> An HTML attachment was scrubbed... >> URL: http://developer.marklogic.com/pipermail/general/ >> attachments/20161111/e6ee38f5/attachment-0001.html >> >> ------------------------------ >> >> Message: 2 >> Date: Fri, 11 Nov 2016 14:52:59 +0100 >> From: Yair Etziony <[email protected]> >> Subject: [MarkLogic Dev General] (no subject) >> To: [email protected] >> Message-ID: >> <CALFBOvH0ggd_ATA8mgctVRHFbKPmMC996H639mGh=P >> [email protected]> >> Content-Type: text/plain; charset="utf-8" >> >> Hi All >> i have a question about ML, this more of DBA or admin question, but i will >> take my shot here. >> >> We run Marklogic on Linux server (Centos 7), there are times when we run >> big processes and the ML daemon would take all memory, in situations like >> that the Linux kernel will kill the process and this is not something we >> can accept in our prod stack. >> the question is what would be the best practice to limit the memory and >> cpu consumption for ML daemon? ( for example using linux cgroups) >> >> Thanks in advance for any answer >> >> bests >> Y >> >> -- >> >> --- >> >> Yair Etziony >> >> LambdaWerk GmbH >> >> Oranienburger Stra?e 87/89 >> >> 10178 Berlin >> >> Phone +49 151 75938044 >> -------------- next part -------------- >> An HTML attachment was scrubbed... >> URL: http://developer.marklogic.com/pipermail/general/ >> attachments/20161111/85f6cb1c/attachment-0001.html >> >> ------------------------------ >> >> _______________________________________________ >> General mailing list >> [email protected] >> Manage your subscription at: >> http://developer.marklogic.com/mailman/listinfo/general >> >> >> End of General Digest, Vol 149, Issue 7 >> *************************************** >> > > > > -- > > Josh Bench > System Administrator > > [email protected] > 888-659-4069 ext.# 897 > > > P.O. Box 44403 | Madison, WI 53744 > WealthCounsel.com > <https://www.facebook.com/WealthCounsel/?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature> > > <https://plus.google.com/u/0/110019845109170387394/posts?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature> > > <https://www.youtube.com/user/WealthCounsel?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature> > > <https://twitter.com/WealthCounsel/?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature> > > <https://www.linkedin.com/company/96425?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature> > > Share the Wealth > <https://member.wealthcounsel.com/referrals?utm_source=WiseStamp&utm_medium=email&utm_term=&utm_content=&utm_campaign=signature>! > Refer a colleague to WealthCounsel and we?ll credit you both one month of > membership dues. > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://developer.marklogic.com/pipermail/general/attachments/20161111/becaa80f/attachment-0001.html > > > ------------------------------ > > Message: 2 > Date: Fri, 11 Nov 2016 15:52:29 -0800 > From: Danny Sokolsky <[email protected]> > Subject: Re: [MarkLogic Dev General] A question about limiting > Marklogic memory and CPU usage. > To: MarkLogic Developer Discussion <[email protected]> > Message-ID: > <cahtrw-9okahb4jbnrfo3ywtvfuvkhtz3xupx_mtbmtmkruw...@mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Make sure you have configured swap space appropriately: > > > https://docs.marklogic.com/guide/installation/intro#id_11335 > > > -Danny > > On Fri, Nov 11, 2016 at 5:32 AM, Yair Etziony <[email protected]> > wrote: > >> Hi All >> i have a question about ML, this more of DBA or admin question, but i will >> take my shot here. >> >> We run Marklogic on Linux server (Centos 7), there are times when we run >> big processes and the ML daemon would take all memory, in situations like >> that the Linux kernel will kill the process and this is not something we >> can accept in our prod stack. >> the question is what would be the best practice to limit the memory and >> cpu consumption for ML daemon? ( for example using linux cgroups) >> >> Thanks in advance for any answer >> >> bests >> Y >> >> -- >> >> --- >> >> Yair Etziony >> >> LambdaWerk GmbH >> >> Oranienburger Stra?e 87/89 >> >> 10178 Berlin >> >> Phone +49 151 75938044 >> >> _______________________________________________ >> General mailing list >> [email protected] >> Manage your subscription at: >> http://developer.marklogic.com/mailman/listinfo/general >> >> > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: > http://developer.marklogic.com/pipermail/general/attachments/20161111/2257289a/attachment-0001.html > > > ------------------------------ > > _______________________________________________ > General mailing list > [email protected] > Manage your subscription at: > http://developer.marklogic.com/mailman/listinfo/general > > > End of General Digest, Vol 149, Issue 8 > *************************************** _______________________________________________ General mailing list [email protected] Manage your subscription at: http://developer.marklogic.com/mailman/listinfo/general
