From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <6e35c0620612241700h50fd3b0aw1d326216c9ae2c3b@mail.gmail.com>
Date: Sun, 24 Dec 2006 17:00:08 -0800
From: "Jack Johnson" <knapjack@gmail.com>
To: "Fans of the OS Plan 9 from Bell Labs" <9fans@cse.psu.edu>
Subject: Re: [9fans] why I'm messing with python: mercurial
In-Reply-To: <69898e3538bf13a71cf2403dd5e3d48e@hamnavoe.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
References: <13426df10612232126p669bd6eaq1411e0546c5dcd7@mail.gmail.com>
	<69898e3538bf13a71cf2403dd5e3d48e@hamnavoe.com>
Topicbox-Message-UUID: f9cf1722-ead1-11e9-9d60-3106f5b1d025

On 12/24/06, Richard Miller <9fans@hamnavoe.com> wrote:
> I've run it on a two-processor VIA vt310.  An extra cpu does help with
> responsiveness.

Doesn't surprise me:

http://www.cs.wisc.edu/~kchak/papers/spin-pact06.pdf

If I skimmed this correctly, one thing I gleaned is that any time you
have more virtual CPUs than real CPUs (or cores, for you hyperthreaded
folks), you're going to take a measurable hit (duh).  Oh, plus one
core/CPU, for the host OS.

So, your Xen setup on a dual-core chip should be very nice.

I also just came across:

http://www.hpl.hp.com/personal/Lucy_Cherkasova/papers/xenqos-mware-camera-ready.pdf

but haven't skimmed it yet.

-Jack