Date: Sat, 3 Nov 2012 08:44:55 -1000 (HST) From: Jeff Roberson <jroberson@jroberson.net> To: "O. Hartmann" <ohartman@zedat.fu-berlin.de> Cc: freebsd-current@freebsd.org Subject: Re: DragonFly vs FreeBSD scheduler Message-ID: <alpine.BSF.2.00.1211030842430.1947@desktop> In-Reply-To: <50953A1B.5050401@zedat.fu-berlin.de> References: <CANuCnH9b20NxDpmzPdE2tdEFCTx7a4u_TvvEJNc0J4yCbDm%2Biw@mail.gmail.com> <20121103091752.0000797f@unknown> <50953A1B.5050401@zedat.fu-berlin.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 3 Nov 2012, O. Hartmann wrote: > Am 11/03/12 15:17, schrieb Mark Felder: >> On Sat, 3 Nov 2012 21:18:55 +0800 >> Alie Tan <alie@affle.com> wrote: >> >>> Hi, >>> >>> No offence, just curious about scheduler and its functionality. >>> >>> What is the different between this two that makes FreeBSD performance far >>> behind DragonFly BSD? http://www.dragonflybsd.org/release32/ >>> >> >> I don't have any details but I do know that Dragonfly has been putting a lot of work into their scheduler. Hopefully some of that will trickle back our way. > > > Obviously they made the right decissions, but a single benchmark with a > DB server like postgresql doesn't tell the whole story. Let's see what > Phoronix will come up with. I'd like to see some more benchmarks of > DragonFly 3.2. > > I doubt that the DragonFly scheduler approaches will go/flow easily into > FreeBSD. But I'd like to see it, even dumping ULE for a better approach. It's not the scheduler. It's lock contention in the vm and buffer cache. The scheduler can only schedule what is runnable. We are working to address this problem. Thanks, Jeff > > > > > > > >
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?alpine.BSF.2.00.1211030842430.1947>