Date: Tue, 16 Oct 2007 22:15:46 +0200 From: "[LoN]Kamikaze" <LoN_Kamikaze@gmx.de> To: josh.carroll@gmail.com Cc: Kris Kennaway <kris@freebsd.org>, freebsd-stable@freebsd.org Subject: Re: SCHED_4BSD in RELENG_7 disturbs workflow Message-ID: <47151BF2.6020103@gmx.de> In-Reply-To: <8cb6106e0710161309o4658f41fse686b637d96be7f1@mail.gmail.com> References: <47150D87.3070804@gmx.de> <47150F82.9060805@FreeBSD.org> <8cb6106e0710161309o4658f41fse686b637d96be7f1@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Josh Carroll wrote: >> Not to say that any problems that might have developed with SCHED_4BSD >> should not be fixed, but you should give SCHED_ULE a try since it brings >> benefits even for single CPU systems (e.g. better interactive response). > > For my particular work load, 4BSD is actually faster than ULE in > RELENG_7. Specifically, on a Q6600 running ffmpeg -threads 8 to > transcode some H.264 video, 4BSD is about 5% faster. I took a sample > video and transcoded the first 120 seconds of it, and here are the > results (including a control from 6.2-RELEASE-p7/4BSD scheduler): > > releng_6_2 (4BSD) 1:32.39 > releng_7 (4BSD) 1:32.44 > releng_7 (ULE) 1:37.15 > > This is obviously a different scenario from MySQL. So perhaps ULE > isn't as well tuned for cases like ffmpeg? > > Josh > I suspect that the increased performance is stolen from the process with focus. I don't care much about my calculation being 5% faster if that renders the machine unusable.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?47151BF2.6020103>