<p>or proportional rate reduction - even better?</p>
<p>just random thoughts..</p>
<div class="gmail_quote">On Oct 27, 2012 12:10 PM, "Arjuna Sathiaseelan" <<a href="mailto:arjuna.sathiaseelan@gmail.com">arjuna.sathiaseelan@gmail.com</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<p>this sounds great actually..probably reduced to 3/4th cwnd on ecn marks?</p>
<p>arjuna</p>
<div class="gmail_quote">On Oct 27, 2012 12:06 PM, <<a href="mailto:iccrg-request@cs.ucl.ac.uk" target="_blank">iccrg-request@cs.ucl.ac.uk</a>> wrote:<br type="attribution"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Send Iccrg mailing list submissions to<br>
<a href="mailto:iccrg@cs.ucl.ac.uk" target="_blank">iccrg@cs.ucl.ac.uk</a><br>
<br>
To subscribe or unsubscribe via the World Wide Web, visit<br>
<a href="http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg" target="_blank">http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg</a><br>
or, via email, send a message with subject or body 'help' to<br>
<a href="mailto:iccrg-request@cs.ucl.ac.uk" target="_blank">iccrg-request@cs.ucl.ac.uk</a><br>
<br>
You can reach the person managing the list at<br>
<a href="mailto:iccrg-owner@cs.ucl.ac.uk" target="_blank">iccrg-owner@cs.ucl.ac.uk</a><br>
<br>
When replying, please edit your Subject line so it is more specific<br>
than "Re: Contents of Iccrg digest..."<br>
<br>
<br>
Today's Topics:<br>
<br>
1. Agenda (Michael Welzl)<br>
2. Why don't we stop treating ECN and loss similarly? (Michael Welzl)<br>
<br>
<br>
----------------------------------------------------------------------<br>
<br>
Message: 1<br>
Date: Sat, 27 Oct 2012 10:20:13 +0200<br>
From: Michael Welzl <<a href="mailto:michawe@ifi.uio.no" target="_blank">michawe@ifi.uio.no</a>><br>
Subject: [Iccrg] Agenda<br>
To: iccrg list <<a href="mailto:iccrg@cs.ucl.ac.uk" target="_blank">iccrg@cs.ucl.ac.uk</a>><br>
Message-ID: <<a href="mailto:9C0FF8B7-8A75-4599-9980-BBA7B76842F0@ifi.uio.no" target="_blank">9C0FF8B7-8A75-4599-9980-BBA7B76842F0@ifi.uio.no</a>><br>
Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes<br>
<br>
Hi,<br>
<br>
There is still space on our agenda - currently, it looks like this:<br>
<br>
<br>
ICCRG agenda, IETF 85, Atlanta, Monday, 5 November, 13:00-15:00, Salon A<br>
<br>
Rong Pan: "A new algorithm for dealing with buffer bloat", 20+10 min<br>
<br>
Nestor Michael C. Tiglao: "Transport layer caching mechanisms and<br>
optimization", 20+10 min<br>
<br>
Open discussion about the applicability of more precise ECN signalling<br>
<br>
<br>
Cheers,<br>
Michael<br>
<br>
<br>
<br>
<br>
------------------------------<br>
<br>
Message: 2<br>
Date: Sat, 27 Oct 2012 11:26:59 +0200<br>
From: Michael Welzl <<a href="mailto:michawe@ifi.uio.no" target="_blank">michawe@ifi.uio.no</a>><br>
Subject: [Iccrg] Why don't we stop treating ECN and loss similarly?<br>
To: iccrg list <<a href="mailto:iccrg@cs.ucl.ac.uk" target="_blank">iccrg@cs.ucl.ac.uk</a>><br>
Message-ID: <<a href="mailto:920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no" target="_blank">920D98EE-EFEB-47E2-879C-84999F258771@ifi.uio.no</a>><br>
Content-Type: text/plain; charset=US-ASCII; format=flowed; delsp=yes<br>
<br>
Hi,<br>
<br>
Here's an idea, inspired by something Bob Briscoe posted to the TSVWG<br>
list recently in a discussion of draft-carlberg-tsvwg-ecn-reactions.<br>
However, this possibly stupid idea is my own responsibility alone :-)<br>
<br>
<br>
According to RFC 3168, senders must react to ECN just as if packets<br>
had been dropped. This is to maintain fairness between ECN-compatible<br>
and non-compatible flows.<br>
Because of this requirement, AQMs cannot ECN-mark packets more<br>
aggressively than it drops packets from non-ECN-capable flows - else<br>
ECN-marked flows would be at a disadvantage.<br>
<br>
We have seen various non-standard congestion control behaviors can co-<br>
exist reasonably well with standard TCP in practice. If it was<br>
possible to have a milder congestion reaction to ECN-based reaction,<br>
it would also be possible to ECN-mark packets earlier, leading to a<br>
bigger advantage for everyone using ECN. And none of this is possible<br>
when we have the "treat an ECN mark just like loss" rule in place.<br>
<br>
Hence, my question: to incentivize ECN usage and enable better<br>
behavior when it's used, shouldn't we remove this rule?<br>
<br>
Note that this is not even about a more fine-grain interpretation of<br>
ECN feedback - it's more like an intermediate step.<br>
<br>
I'm curious what everyone thinks... am I missing something?<br>
<br>
Cheers,<br>
Michael<br>
<br>
<br>
<br>
<br>
------------------------------<br>
<br>
_______________________________________________<br>
Iccrg mailing list<br>
<a href="mailto:Iccrg@cs.ucl.ac.uk" target="_blank">Iccrg@cs.ucl.ac.uk</a><br>
<a href="http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg" target="_blank">http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg</a><br>
<br>
<br>
End of Iccrg Digest, Vol 81, Issue 8<br>
************************************<br>
</blockquote></div>
</blockquote></div>