Re: [mif] Happy Eyeballs Extension for MIF

GangChen <phdgang@gmail.com> Tue, 05 April 2011 15:17 UTC

Return-Path: <phdgang@gmail.com>
X-Original-To: mif@core3.amsl.com
Delivered-To: mif@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 2480D28C0EC for <mif@core3.amsl.com>; Tue, 5 Apr 2011 08:17:03 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.999
X-Spam-Level:
X-Spam-Status: No, score=-2.999 tagged_above=-999 required=5 tests=[AWL=-0.400, BAYES_00=-2.599]
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 wEN3+Sxdvdrk for <mif@core3.amsl.com>; Tue, 5 Apr 2011 08:17:02 -0700 (PDT)
Received: from mail-vx0-f172.google.com (mail-vx0-f172.google.com [209.85.220.172]) by core3.amsl.com (Postfix) with ESMTP id 6CB233A6949 for <mif@ietf.org>; Tue, 5 Apr 2011 08:17:02 -0700 (PDT)
Received: by vxg33 with SMTP id 33so442474vxg.31 for <mif@ietf.org>; Tue, 05 Apr 2011 08:18:45 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:in-reply-to:references:date :message-id:subject:from:to:cc:content-type; bh=Tp0HDrbSVqpB1ZW2IaBTp2dWPclV9vN1KOku+ga04mo=; b=t1mO20miJBWdARx9rQuMsu9gRBtI+Cr6VcmMW90MdRL9HExJAQnUh0IyCEop6rCvfV vKsTVK1rNk6peVFV7Q1h3JrbMi7gR5UhCqJCOiEJoteN4rQWx83p1gJ+ow5W+ZeVO9fs hsRgy+qRlof+F8KKnJAEHFWoF6f5UgEPcXIgs=
DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; b=Qq3InWc0x49pUWleN8WgeikJX46R4HfQUHSOkWztUZPvh4224mF1iA4c5SYtyHEScj WOzF31vk/RBDDFEHiFhD2mlLOnKbmMfmab8k8EuRVQCkoZIIfh+QJBG1qPb5Dek1Yz2X D9xjgvZ/qTHACxszLwdnShtT5BNvIUJrj4mfQ=
MIME-Version: 1.0
Received: by 10.52.166.39 with SMTP id zd7mr7066517vdb.311.1302016725450; Tue, 05 Apr 2011 08:18:45 -0700 (PDT)
Received: by 10.52.164.132 with HTTP; Tue, 5 Apr 2011 08:18:45 -0700 (PDT)
In-Reply-To: <916CE6CF87173740BC8A2CE443096962014D2E@008-AM1MPN1-036.mgdnok.nokia.com>
References: <AANLkTim+jfEdXfkbrYxYYm5jeeT_fsMvpeV+fZYDN0rY@mail.gmail.com> <916CE6CF87173740BC8A2CE443096962014D2E@008-AM1MPN1-036.mgdnok.nokia.com>
Date: Tue, 05 Apr 2011 23:18:45 +0800
Message-ID: <BANLkTinvr6efQ_DLZBYg6Qhk3wAfXcwNRg@mail.gmail.com>
From: GangChen <phdgang@gmail.com>
To: teemu.savolainen@nokia.com
Content-Type: text/plain; charset="ISO-8859-1"
Cc: mif@ietf.org
Subject: Re: [mif] Happy Eyeballs Extension for MIF
X-BeenThere: mif@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: Multiple Interface Discussion List <mif.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/mif>, <mailto:mif-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/mif>
List-Post: <mailto:mif@ietf.org>
List-Help: <mailto:mif-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/mif>, <mailto:mif-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 05 Apr 2011 15:17:03 -0000

Hello Teemu,

Please see reply in line.
Some thoughts updates have been included according to the meeting discussion.


2011/3/27, teemu.savolainen@nokia.com <teemu.savolainen@nokia.com>:
> Hi,
>
> This is interesting work. Couple of questions:
>
> Is the value of I initially zero for any destination/hostname that has not
> yet been connected to?

[Gang] The value of I is supposed to be zero initially in preliminary
algorithm proposal.

> draft-ietf-mif-dhcpv6-route-option and RFC4191 allow definition of specific
> routes. In the presence of specific routes it sounds, for me at least,
> somewhat odd to send connection attempts via all interfaces. What about
> incrementing I for a destination address that matches a more specific route?
> Hence initial connection attempt would be sent over the interface that has
> matching route and other interfaces would be tried only if no reply on the
> preferred one?

[Gang] We could take routing information into account for computing value I.
Also, I have noticed there are several preconditions impacting on
value I, like interface cost, higher layer service demanding, etc.
All preconditions would contribute to value I computations. I guess we
need to formulate a function to reflect all criteria.
Multiple rules make algorithm complicated. I'm still working on
modeling the different interface behavior based on several
preconditions.
Any suggestion about this?


> How does the value of I relate to time? (i.e. if I is -1 does the node wait
> 10ms before trying corresponding less-preferred interface).

[Gang] How about absolute value of I*10 milliseconds which is in line
with original Happy Eyeballs proposal

> Why would the DNS query be sent over all interfaces? Is it because "I" for a
> destination name is initially a zero? Like in routing case, if Improved DNS
> Server Selection is in play, would the "I" be nonzero for those interfaces
> over which DNS suffix matching the requested name has been received on?

[Gang] I think that has same situation with second comments.

Best Regards

Gang