From nobody Fri Jul 10 04:10:55 2015 Return-Path: X-Original-To: dclc@ietfa.amsl.com Delivered-To: dclc@ietfa.amsl.com Received: from localhost (ietfa.amsl.com [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 290701A1BF8 for ; Fri, 10 Jul 2015 04:10:54 -0700 (PDT) X-Virus-Scanned: amavisd-new at amsl.com X-Spam-Flag: NO X-Spam-Score: -1.6 X-Spam-Level: X-Spam-Status: No, score=-1.6 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, MIME_8BIT_HEADER=0.3, SPF_PASS=-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 zRJH4b-XsN-w for ; Fri, 10 Jul 2015 04:10:52 -0700 (PDT) Received: from server.dnsblock1.com (server.dnsblock1.com [85.13.236.178]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 9AB7F1A9045 for ; Fri, 10 Jul 2015 04:10:52 -0700 (PDT) Received: from 151.215.199.146.dyn.plus.net ([146.199.215.151]:44097 helo=[192.168.0.7]) by server.dnsblock1.com with esmtpsa (TLSv1.2:DHE-RSA-AES128-SHA:128) (Exim 4.85) (envelope-from ) id 1ZDWCg-0003qx-EE; Fri, 10 Jul 2015 12:10:50 +0100 Message-ID: <559FA839.80002@bobbriscoe.net> Date: Fri, 10 Jul 2015 12:10:49 +0100 From: Bob Briscoe User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.7.0 MIME-Version: 1.0 To: =?UTF-8?B?6YKT54G16I6JL0xpbmdsaSBEZW5n?= , Fred Baker References: <559F9E69.5020804@bobbriscoe.net> In-Reply-To: <559F9E69.5020804@bobbriscoe.net> Content-Type: multipart/alternative; boundary="------------090000010103010006010407" X-AntiAbuse: This header was added to track abuse, please include it with any abuse report X-AntiAbuse: Primary Hostname - server.dnsblock1.com X-AntiAbuse: Original Domain - irtf.org X-AntiAbuse: Originator/Caller UID/GID - [47 12] / [47 12] X-AntiAbuse: Sender Address Domain - bobbriscoe.net X-Get-Message-Sender-Via: server.dnsblock1.com: authenticated_id: in@bobbriscoe.net Archived-At: Cc: "De Schepper, Koen \(Koen\)" , dclc@irtf.org Subject: [Dclc] Data Centre to the Home: Awesome Low Latency, Low Loss and Scalable Throughput X-BeenThere: dclc@irtf.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: Discussion of Data Center Latency Control List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 10 Jul 2015 11:10:54 -0000 This is a multi-part message in MIME format. --------------090000010103010006010407 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit Lingli, Fred, DCLC list, FYI - see links in posting below. Highlights: * 1ms 99%ile queuing delay for all DCTCP traffic over an e2e test network with real broadband equipment. * DCTCP co-existence with Reno & Cubic, with no transport ID inspection. * ~20 lines of code. * Zero config As well as public Internet, this is intended for heterogeneous multi-administrator DCTCP/Reno/Cubic data centres. You will notice that this new AQM works with DCTCP 'as is' but some 'safety features' must be added to DCTCP (such as fall-back to Reno or Cubic on loss). We plan to work with those maintaining the DCTCP code to make these changes. Bob PS. Note my new interim email while I am between jobs. -------- Forwarded Message -------- Subject: Awesome Low Latency, Low Loss and Scalable Throughput Date: Fri, 10 Jul 2015 11:28:57 +0100 From: Bob Briscoe To: Scheffenegger, Richard , Wesley Eddy , AQM IETF list RIchard, Wes, AQM List, We've been working as fast as we can doing thousands of experiments over an e2e test network with real broadband equipment to check that the awesomeness we are seeing with this new AQM is valid in many different traffic scenarios. Even at very high loads, we are still getting 99th %ile queuing latency of about 1ms. We tried to have an I-D written-up for Monday's deadline, but just missed it. I have uploaded it to my personal Web site here and we will post it to the IETF servers once they re-open next Monday: We also have a paper with plots of (a small selection of) all the experiments. It is under submission to a conference, but we have uploaded a pre-print to my site too: ‘Data Centre to the Home’: Ultra-Low Latency for All Koen will be presenting this in the AQM session in Prague (some of the details might be bumped to a supplementary slot in ICCRG), but we wanted to give you as much material as possible in advance. Cheers Bob -- ________________________________________________________________ Bob Briscoehttp://bobbriscoe.net/ -- ________________________________________________________________ Bob Briscoe http://bobbriscoe.net/ --------------090000010103010006010407 Content-Type: text/html; charset=utf-8 Content-Transfer-Encoding: 8bit Lingli, Fred, DCLC list,

FYI - see links in posting below.

Highlights:
* 1ms 99%ile queuing delay for all DCTCP traffic over an e2e test network with real broadband equipment.
* DCTCP co-existence with Reno & Cubic, with no transport ID inspection.
* ~20 lines of code.
* Zero config

As well as public Internet, this is intended for heterogeneous multi-administrator DCTCP/Reno/Cubic data centres.

You will notice that this new AQM works with DCTCP 'as is' but some 'safety features' must be added to DCTCP (such as fall-back to Reno or Cubic on loss). We plan to work with those maintaining the DCTCP code to make these changes.



Bob

PS. Note my new interim email while I am between jobs.

-------- Forwarded Message --------
Subject: Awesome Low Latency, Low Loss and Scalable Throughput
Date: Fri, 10 Jul 2015 11:28:57 +0100
From: Bob Briscoe <ietf@bobbriscoe.net>
To: Scheffenegger, Richard <rs@netapp.com>, Wesley Eddy <wes@mti-systems.com>, AQM IETF list <aqm@ietf.org> 


RIchard, Wes, AQM List,

We've been working as fast as we can doing thousands of experiments over 
an e2e test network with real broadband equipment to check that the 
awesomeness we are seeing with this new AQM is valid in many different 
traffic scenarios.

Even at very high loads, we are still getting 99th %ile queuing latency 
of about 1ms.

We tried to have an I-D written-up for Monday's deadline, but just 
missed it. I have uploaded it to my personal Web site here and we will 
post it to the IETF servers once they re-open next Monday:
<http://www.bobbriscoe.net/projects/latency/draft-briscoe-aqm-dualq-coupled-00.txt>

We also have a paper with plots of (a small selection of) all the 
experiments. It is under submission to a conference, but we have 
uploaded a pre-print to my site too:
‘Data Centre to the Home’: Ultra-Low Latency for All
<http://www.bobbriscoe.net/projects/latency/dctth_preprint.pdf>

Koen will be presenting this in the AQM session in Prague (some of the 
details might be bumped to a supplementary slot in ICCRG), but we wanted 
to give you as much material as possible in advance.

Cheers


Bob

-- 
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/





-- 
________________________________________________________________
Bob Briscoe                               http://bobbriscoe.net/
--------------090000010103010006010407--