Re: [alto] How Data Center Virtualization influence ALTO mechanism.

Enrico Marocco <enrico.marocco@telecomitalia.it> Tue, 12 October 2010 08:24 UTC

Return-Path: <enrico.marocco@telecomitalia.it>
X-Original-To: alto@core3.amsl.com
Delivered-To: alto@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 795333A6BFB for <alto@core3.amsl.com>; Tue, 12 Oct 2010 01:24:56 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -99.234
X-Spam-Level:
X-Spam-Status: No, score=-99.234 tagged_above=-999 required=5 tests=[AWL=-0.004, BAYES_05=-1.11, HELO_EQ_IT=0.635, HOST_EQ_IT=1.245, USER_IN_WHITELIST=-100]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id eAepCi3ehzW8 for <alto@core3.amsl.com>; Tue, 12 Oct 2010 01:24:55 -0700 (PDT)
Received: from GRFEDG701BA020.telecomitalia.it (grfedg701ba020.telecomitalia.it [156.54.233.200]) by core3.amsl.com (Postfix) with ESMTP id B9AE53A67A3 for <alto@ietf.org>; Tue, 12 Oct 2010 01:24:53 -0700 (PDT)
Received: from GRFHUB701BA020.griffon.local (10.188.101.111) by GRFEDG701BA020.telecomitalia.it (10.188.45.100) with Microsoft SMTP Server (TLS) id 8.2.254.0; Tue, 12 Oct 2010 10:26:05 +0200
Received: from MacLab.local (163.162.173.7) by smtp.telecomitalia.it (10.188.101.114) with Microsoft SMTP Server (TLS) id 8.2.254.0; Tue, 12 Oct 2010 10:26:04 +0200
Message-ID: <4CB41B9B.5090609@telecomitalia.it>
Date: Tue, 12 Oct 2010 10:26:03 +0200
From: Enrico Marocco <enrico.marocco@telecomitalia.it>
User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.9) Gecko/20100915 Thunderbird/3.1.4
MIME-Version: 1.0
To: Lars Eggert <lars.eggert@nokia.com>
References: <006201cb6782$2cb463a0$45548a0a@china.huawei.com> <C8E8F68D-1560-4329-B370-D5A182C7F6F5@nokia.com>
In-Reply-To: <C8E8F68D-1560-4329-B370-D5A182C7F6F5@nokia.com>
Content-Type: multipart/signed; protocol="application/pkcs7-signature"; micalg="sha1"; boundary="------------ms010303050706020601050709"
Cc: "alto@ietf.org" <alto@ietf.org>
Subject: Re: [alto] How Data Center Virtualization influence ALTO mechanism.
X-BeenThere: alto@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: "Application-Layer Traffic Optimization \(alto\) WG mailing list" <alto.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/alto>, <mailto:alto-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/alto>
List-Post: <mailto:alto@ietf.org>
List-Help: <mailto:alto-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/alto>, <mailto:alto-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 12 Oct 2010 08:24:56 -0000

On 10/12/10 9:17 AM, Lars Eggert wrote:
>> IP-based, Vms are closer to DC-A. Physically, these VMs are much
>> closer to hosts in DC-B. However things are not so easy, especially
>> considering how these VMs are routed. Current ALTO may give wrong
>> cost ranking.
> 
> If you're moving VMs around in the topology, you'll have to update
> the announced ALTO information as you do that.

According to a presentation I saw last year, this is actually what
happens in Facebook's server farms, where BitTorrent and a customized
locality-aware tracker are used to distribute software updates
minimizing inter-rack traffic. Whenever the topology changes -- and I
guess that happens quite often -- the updated information is directly
loaded into the tracker. I don't remember how the update mechanism works
in that specific case, but I guess that could be quite effectively
automated with some kind of integration between the ALTO-like service
and some IGP database.

I can't find a pointer to that presentation, but an article and a video
explaining how software update works in Facebook is available on
TorrentFreak:
http://torrentfreak.com/facebook-uses-bittorrent-and-they-love-it-100625/

-- 
Ciao,
Enrico