Re: [Roll] [roll] #157 (draft-doi-roll-mpl-parameter-configuration): mpl-parameter-configuration-00 - Effect of inconsistent parameter set among nodes

peter van der Stok <stokcons@xs4all.nl> Fri, 01 August 2014 08:13 UTC

Return-Path: <stokcons@xs4all.nl>
X-Original-To: roll@ietfa.amsl.com
Delivered-To: roll@ietfa.amsl.com
Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 98B0B1A045B for <roll@ietfa.amsl.com>; Fri, 1 Aug 2014 01:13:25 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: 2.693
X-Spam-Level: **
X-Spam-Status: No, score=2.693 tagged_above=-999 required=5 tests=[BAYES_20=-0.001, HELO_EQ_NL=0.55, HOST_EQ_NL=1.545, J_CHICKENPOX_41=0.6, RCVD_IN_DNSWL_NONE=-0.0001, RP_MATCHES_RCVD=-0.001] autolearn=no
Received: from mail.ietf.org ([4.31.198.44]) by localhost (ietfa.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id VPjQViMDzLPv for <roll@ietfa.amsl.com>; Fri, 1 Aug 2014 01:13:23 -0700 (PDT)
Received: from smtp-vbr5.xs4all.nl (smtp-vbr5.xs4all.nl [194.109.24.25]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 181271A036F for <roll@ietf.org>; Fri, 1 Aug 2014 01:13:22 -0700 (PDT)
Received: from roundcube.xs4all.nl (roundcube3.xs4all.net [194.109.20.199]) by smtp-vbr5.xs4all.nl (8.13.8/8.13.8) with ESMTP id s718DB7W018376; Fri, 1 Aug 2014 10:13:12 +0200 (CEST) (envelope-from stokcons@xs4all.nl)
Received: from a82-95-140-48.adsl.xs4all.nl ([82.95.140.48]) by roundcube.xs4all.nl with HTTP (HTTP/1.1 POST); Fri, 01 Aug 2014 10:13:11 +0200
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"; format="flowed"
Content-Transfer-Encoding: 7bit
Date: Fri, 01 Aug 2014 10:13:11 +0200
From: peter van der Stok <stokcons@xs4all.nl>
To: Routing Over Low power and Lossy networks <roll@ietf.org>
Organization: vanderstok consultancy
Mail-Reply-To: consultancy@vanderstok.org
In-Reply-To: <3398D003-F294-455D-A036-AB7C92CDB1E8@cs.stanford.edu>
References: <067.f0fd62a4161a1ac153a9b089b2780010@trac.tools.ietf.org> <538302CC.9020904@toshiba.co.jp> <3398D003-F294-455D-A036-AB7C92CDB1E8@cs.stanford.edu>
Message-ID: <01206ae54ce5202d5afdc5225bb5653d@xs4all.nl>
X-Sender: stokcons@xs4all.nl (Pi7IaY9fn0RyWGcpSYkR/bhMQJS5NGgv)
User-Agent: XS4ALL Webmail
X-Virus-Scanned: by XS4ALL Virus Scanner
Archived-At: http://mailarchive.ietf.org/arch/msg/roll/o6wX3DFzf01jIXYoFrwuoTBMZMI
Cc: draft-doi-roll-mpl-parameter-configuration@tools.ietf.org
Subject: Re: [Roll] [roll] #157 (draft-doi-roll-mpl-parameter-configuration): mpl-parameter-configuration-00 - Effect of inconsistent parameter set among nodes
X-BeenThere: roll@ietf.org
X-Mailman-Version: 2.1.15
Precedence: list
Reply-To: consultancy@vanderstok.org, Routing Over Low power and Lossy networks <roll@ietf.org>
List-Id: Routing Over Low power and Lossy networks <roll.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/roll>, <mailto:roll-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www.ietf.org/mail-archive/web/roll/>
List-Post: <mailto:roll@ietf.org>
List-Help: <mailto:roll-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/roll>, <mailto:roll-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 01 Aug 2014 08:13:25 -0000

> Depending on K, a difference in Imax can do more than just
> increase/reduce responsibility. It could actually lead to complete
> suppression.

Is it not better to consider the relation between K and the number of 
1-hop MPL neighbours?
Supposing M is the number of 1-hop neighbours of a node, one can reason 
about:
K > M, where no suppression occurs
K < M there is an increased probability of suppression that increases 
with increasing Imax

Greetings,

Peter

Philip Levis schreef op 2014-07-29 19:25:
> On May 26, 2014, at 2:01 AM, Yusuke DOI <yusuke.doi@toshiba.co.jp> 
> wrote:
> 
>> Hi,
>> 
>> Thank you for summarizing issues on my draft.
>> On #157, my understanding is as follows. If there are any issues you 
>> know, please let me know.
>> 
>> Inconsistent but valid parameter does not harm the network much (it 
>> may harm battery-powered nodes).
> 
> Given the L in RPL/MPL includes the term "low power", harming
> battery-powered nodes is kind of important. It's easy to try to ignore
> them because they're so challenging, but I'd strongly recommend
> against it.
> 
>> 
>> Reason:
>> 
>> 1) If a node have smaller Imin/Imax, it takes more responsibility to 
>> repeat messages than surrounding nodes. The node will consume more 
>> power and the area will have higher traffic than expected until MPL 
>> parameters of the node is updated.
> 
> It's worth noting that transmission (in most link layers MPL/RPL care
> about) does not consume significantly more energy than reception.
> Therefore, one node that has a smaller Imin or Imax introduces an
> energy cost on *all* of the nodes around it.
> 
> A smaller Imin can double the latency of propagation over a single
> hop, as nodes, on receiving an update, receive multiple Trickle
> messages and so suppress for their first interval.
> 
> Depending on K, a difference in Imax can do more than just
> increase/reduce responsibility. It could actually lead to complete
> suppression. Consider the case where three nodes have Imax=t and one
> node has Imax=t+2 (so the actual time interval is 4 times greater).
> K<=3. The three nodes will almost always completely suppress the
> fourth node with the longer interval.
> 
>> 2) If a node have larger Imin/Imax, it takes less responsibility to 
>> send messages than surrounding nodes. As other nodes will send the 
>> message instead of the node with old parameters, effect for traffic 
>> load and e2e delay is negligible on dense mesh clusters. If the node 
>> with old parameters is on sparse area of a mesh network, larger 
>> Imin/Imax will cause larger e2e delay than new parameter's expectation 
>> untilthe node is updated.
> 
> You should disambiguate the effects of Imin and Imax.
> 
>> 3) If a node have smaller K, it takes less responsibility to repeat 
>> messages than surrounding nodes. On dense network, the effect is 
>> negligible. On sparse network, messages pass through the node will 
>> have less reliability until the node is updated.
> 
> It also introduces uneven transmission load. A node with a higher K
> will transmit with a much higher probability that those with the lower
> K (within the logarithmic bounds introduced by packet loss).
> 
> I'd suggest re-reading 6206 (Section 6), which goes into greater
> detail than the bullet points above and is much more specific.
> 
>> 4) If a node have larger K, it will repeat almost all the messages. 
>> The area will have excessive traffic untile the node is updated.
>> 
> 
> This is not true. Note that packet loss/imperfect duplicate
> suppression/uneven topologies often cause some nodes to receive more
> than K packets. Again, I'd suggest re-reading 6206.
> 
> Phil
> _______________________________________________
> Roll mailing list
> Roll@ietf.org
> https://www.ietf.org/mailman/listinfo/roll