Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 25 Jan 2012 08:50:54 -0500
From:      John Baldwin <jhb@freebsd.org>
To:        Ivan Voras <ivoras@freebsd.org>
Cc:        svn-src-head@freebsd.org, svn-src-all@freebsd.org, Florian Smeets <flo@freebsd.org>, src-committers@freebsd.org, Andriy Gapon <avg@freebsd.org>
Subject:   Re: svn commit: r228424 - in head/sys: kern sys
Message-ID:  <201201250850.54443.jhb@freebsd.org>
In-Reply-To: <CAF-QHFW_sM7=euOtAUrJpn6e4k1MpWyiD_7Az%2B=SJenm5DLdpQ@mail.gmail.com>
References:  <201112112102.pBBL21kB068967@svn.freebsd.org> <4F1DE4FF.3080606@FreeBSD.org> <CAF-QHFW_sM7=euOtAUrJpn6e4k1MpWyiD_7Az%2B=SJenm5DLdpQ@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday, January 24, 2012 5:08:39 am Ivan Voras wrote:
> On 23 January 2012 23:53, Florian Smeets <flo@freebsd.org> wrote:
> 
> > which creates a database work set of ~1.5GB. Max throughput was achieved
> > at 20 Clients.
> 
> > At 40 threads the results varied between 43000 - 76500 across reboots.
> > Attilio suspects that this can be caused by the kernel memory layout
> > changing under the woods creating cache effects difficult to control,
> > therefor the scaling factor was reduced to 10 (~150MB work set) and the
> > numbers got deterministic across reboot.
> 
> Or possibly NUMA? Though 40 processes and 1.5 GB seem too low for NUMA
> effects to be so noticable...

It can be noticable for at least some workloads.  It may reduce some of the 
noise in this case.

> Was the round-robin allocator talked about in here:
> http://lists.freebsd.org/pipermail/freebsd-hackers/2011-October/036525.html
> ever actually committed? I seem to remember some other thread which
> said it wasn't yet but can't find it now, and I also cannot find the
> commit.

It is not.

> AFAIK the current state of NUMA is still described in
> http://svn.freebsd.org/changeset/base/210550

Yes.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201201250850.54443.jhb>