Date: Mon, 17 May 1999 16:19:13 -0700 From: Julian Elischer <julian@whistle.com> To: Joao Carlos Mendes Luis <jonny@jonny.eng.br> Cc: Matthew Dillon <dillon@apollo.backplane.com>, dg@root.com, hackers@FreeBSD.ORG Subject: Re: Seti project / stats reset, new version available Message-ID: <3740A3F1.446B9B3D@whistle.com> References: <199905172119.SAA14200@roma.coe.ufrj.br>
next in thread | previous in thread | raw e-mail | index | archive | help
Joao Carlos Mendes Luis wrote: > > #define quoting(Matthew Dillon) > // This reduces the effect an idprio seti background task has on the rest of > // the system. > > Yeah... I was going to ask just this. My users complained about > server response after starting setiathome. > > // sysctl -w kern.quantum=20000 > > Humm, my systems are configured with a value with 10: WARNING: on 3.1 or 2.2.8 it has the inverse action to that required.. (braino on my part) so I think it needs to be set to the inverse of20mSec which is 50Hz (i.e. set it to 50 to achieve the same result.. this is confusing but ....) julian > > krakatoa::root [650] sysctl kern | grep quantum > kern.quantum: 10 > > This both in 3.1-stable (Early May) and 2.2.8-stable (latest). > > Are you talking about a 4.0-current setup ? > > // The default is 100,000 ( 100ms ) which, for a modern cpu, is > // much more chunky then it needs to be. Reducing it to 20ms > // makes a big difference. At 100000 I can feel the slow response > // with seti running in the background. At 20000 it is much less > // pronounced. > > What are the drawbacks of this ? More CPU time spent in context > switches ? yeah a TINY amount more. > > Jonny > > -- > João Carlos Mendes Luís jonny@jonny.eng.br > Networking Enginner jcml@ieee.org > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3740A3F1.446B9B3D>