
From nobody Tue Jul  5 11:44:32 2016
Return-Path: <wes@mti-systems.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CE88C12B050 for <aqm@ietfa.amsl.com>; Tue,  5 Jul 2016 11:44:30 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.601
X-Spam-Level: 
X-Spam-Status: No, score=-2.601 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, RCVD_IN_DNSWL_LOW=-0.7, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mti-systems-com.20150623.gappssmtp.com
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 tyaadt4s6Qi8 for <aqm@ietfa.amsl.com>; Tue,  5 Jul 2016 11:44:28 -0700 (PDT)
Received: from mail-io0-x22f.google.com (mail-io0-x22f.google.com [IPv6:2607:f8b0:4001:c06::22f]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 6A08D12D52D for <aqm@ietf.org>; Tue,  5 Jul 2016 11:44:28 -0700 (PDT)
Received: by mail-io0-x22f.google.com with SMTP id t74so181611938ioi.0 for <aqm@ietf.org>; Tue, 05 Jul 2016 11:44:28 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mti-systems-com.20150623.gappssmtp.com; s=20150623; h=subject:to:references:from:message-id:date:user-agent:mime-version :in-reply-to:content-transfer-encoding; bh=pIYHMio0BUGnBB2RM3Qew+lofzG/jCOuhKR7OnBe+4o=; b=xTrYzaZlie7D9xNUjgnRDzew5RZIs1mCQW+ob+LLUHcI5d1stpqStMe2JYyrZA5Y15 LI7l1C31Ziz4HuamMKG/Xga/KU04r20KH4bsfRY6k/OTyT79YKuwzGHa8HvkavuHQZOX r7LhJeYuk/B9jXMdYe47iBaT8LnNkfbywJwFpTc+6KaE6z3lXMtMH3z5o3OL7GD75P0n vt5p2uHkZAB9bNy5n/AoMflmdctdqq3l3jkqKsBWByUlt+pd91xuVOrMet3o4FMrTBlV enM9A894bi0gD7or1PM5IO9IHi2vraW4wxKWLRL31IdqhyO7EKtu52T77kb7/DHl7ecz eZ9Q==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:subject:to:references:from:message-id:date :user-agent:mime-version:in-reply-to:content-transfer-encoding; bh=pIYHMio0BUGnBB2RM3Qew+lofzG/jCOuhKR7OnBe+4o=; b=l/Q67eBBnDVPO0UpVnziVypilFBAxrmVpjZZjqlwCN9o9IV1M3jIlhMRP+xvTtTZAD sLU/2tGfmR1V0Y6E5y225PRGGVhStcKmS4DGQW2E7NN4dfscG+1SX6ZqeX7nvvlYaWPt dNrvs1rgW+z9wZMfzxHVnS2sIoPZn5tGJMCNPq27omkteoXfHWtRF0kG0pgwh1c1oyG7 KeUoyNj6CS39T826geHKhU1sZZ/6dVkpIWFsSfeHqTtgU13hUBPgKcU8QrQZiDoshSu6 envhVNVaW3CTHOqaqLgxm5vwUPxKKmArWfLBBtRrYmAT0gw3rTkt9m7ouqGJ5UyQ12EH HTdg==
X-Gm-Message-State: ALyK8tKZ5TBESAzTjbN/Ayt9rKWxKD8uGU3//ciB+IXdOo3F1LwbQrboNV5bbdm/7U1R+A==
X-Received: by 10.107.16.164 with SMTP id 36mr16846468ioq.124.1467744267517; Tue, 05 Jul 2016 11:44:27 -0700 (PDT)
Received: from [192.168.1.104] (cpe-76-188-215-129.neo.res.rr.com. [76.188.215.129]) by smtp.gmail.com with ESMTPSA id m133sm2171412ioa.23.2016.07.05.11.44.26 for <aqm@ietf.org> (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 05 Jul 2016 11:44:26 -0700 (PDT)
To: "aqm@ietf.org" <aqm@ietf.org>
References: <dbf541b6-dbd9-21ad-d207-f9a1ca8338a1@mti-systems.com>
From: Wesley Eddy <wes@mti-systems.com>
Message-ID: <dae4660d-bfbf-d964-821b-1ed20018c69e@mti-systems.com>
Date: Tue, 5 Jul 2016 14:44:25 -0400
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.1.1
MIME-Version: 1.0
In-Reply-To: <dbf541b6-dbd9-21ad-d207-f9a1ca8338a1@mti-systems.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/m3TV1O76fTuqWv87LyTQLxtlluM>
Subject: Re: [aqm] Berlin meeting
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Tue, 05 Jul 2016 18:44:31 -0000

Thanks to those who provided feedback.

We will plan to go ahead with the face-to-face meeting since several 
people are interested and will post an agenda shortly.

As always, please continue to also discuss topics on the mailing list.  
Several people have shared some thoughts on rechartering topics, though 
it's not clear that there is yet much agreement on specific topics that 
there would be common energy for.



On 6/29/2016 10:06 AM, Wesley Eddy wrote:
> Hello, as you might have noticed, for Berlin, the AQM group is 
> scheduled to meet on Friday: 
> https://datatracker.ietf.org/meeting/96/agenda.html
>
> The main goal of having the in-person meeting is to talk about closing 
> the WG or rechartering.
>
> As this is in the last slot, on the last day, and many people are 
> likely to be trying to return home, we need to make sure there will be 
> a critical mass to meaningfully hold a meeting, and not waste time.  
> We are also having trouble getting both WG chairs present, in addition.
>
> To gauge if it will be productive to meet, or if we should pursue 
> other options, we created a Doodle poll:
>
> http://doodle.com/poll/tht446xwu25uywat
>
>
> Your feedback will be appreciated.
>
> Note that more than one option can be selected.
>
> A more verbose explanation of the options is:
>
> 1) "In-person meeting in Berlin" - this means we go along with the 
> current plan and hold the meeting in Berlin
>
> 2) "Webex meeting in August" - this means the chairs would setup a 
> webex to discuss rechartering in the month following the IETF meeting
>
> 3) "Mailing list only" - means you think we can figure out what to do 
> from the mailing list only
>
> 4) "Close the WG; no interest to discuss rechartering" - means that 
> you don't think we need to talk about this at all, and think the WG 
> can spin down
>
>
> Thanks for your help in this!
>
>


From nobody Thu Jul  7 04:49:53 2016
Return-Path: <polina.goltsman@student.kit.edu>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 6021312D7B1 for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 04:49:50 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -5.625
X-Spam-Level: 
X-Spam-Status: No, score=-5.625 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_MED=-2.3, RP_MATCHES_RCVD=-1.426] autolearn=ham autolearn_force=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 cUCp8_lc0GDE for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 04:49:46 -0700 (PDT)
Received: from scc-mailout-kit-02.scc.kit.edu (scc-mailout-kit-02.scc.kit.edu [IPv6:2a00:1398:9:f712::810d:e752]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id D88C012D0C7 for <aqm@ietf.org>; Thu,  7 Jul 2016 04:49:45 -0700 (PDT)
Received: from kit-msx-28.kit.edu ([2a00:1398:9:f612::28]) by scc-mailout-kit-02.scc.kit.edu with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA384:256) (envelope-from <polina.goltsman@student.kit.edu>) id 1bL7oK-0004cj-UY; Thu, 07 Jul 2016 13:49:43 +0200
Received: from kit-msx-31.kit.edu (2a00:1398:9:f612::31) by KIT-MSX-28.kit.edu (2a00:1398:9:f612::28) with Microsoft SMTP Server (TLS) id 15.0.1178.4; Thu, 7 Jul 2016 13:49:38 +0200
Received: from kit-msx-31.kit.edu ([fe80::281f:cf9a:1283:1d90]) by kit-msx-31.kit.edu ([fe80::281f:cf9a:1283:1d90%16]) with mapi id 15.00.1178.000; Thu, 7 Jul 2016 13:49:38 +0200
From: "Goltsman, Polina" <polina.goltsman@student.kit.edu>
To: Andrew Mcgregor <andrewmcgr@google.com>
Thread-Topic: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
Thread-Index: AQHRfxbvhSr+gsN4dEOPal6Mw2ErA59ibXCAgAIArACAcubuAIA2HfLC
Date: Thu, 7 Jul 2016 11:49:38 +0000
Message-ID: <1467892048445.64588@student.kit.edu>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu>, <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com>
In-Reply-To: <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com>
Accept-Language: en-US, de-DE
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [2a00:1398:2:4006:527b:9dff:fee2:e7f7]
Content-Type: multipart/alternative; boundary="_000_146789204844564588studentkitedu_"
MIME-Version: 1.0
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/ENA1VZmcFVXCJWrMIbmey4wAYnw>
Cc: Wesley Eddy <wes@mti-systems.com>, Jana Iyengar <jri@google.com>, "Bless, Roland \(TM\)" <roland.bless@kit.edu>, "aqm@ietf.org" <aqm@ietf.org>, Dave Taht <dave.taht@gmail.com>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 11:49:50 -0000

--_000_146789204844564588studentkitedu_
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Dear Andrew, Dear Wesley,


First of all, sorry for the long waiting time. I still haven't processed "n=
its" but I have 2 major issues.


Issue 1: This issue is still related to the reentering dropping state.


As of the latest I-D, the "reentering dropping state"  appears in form of o=
ne sentence in the text and a small code block in the pseudo-code.


The text is on page 13:

>Additional logic prevents re-
>entering the dropping state too soon after exiting it and resumes the
>dropping state at a recent control level, if one exists.

The comment to the pseudo-code on page 20:


           // If min went above target close to when it last went
           // below, assume that the drop rate that controlled the
           // queue on the last cycle is a good starting point to
           // control it now. ('drop_next' will be at most 'interval'
           // later than the time of the last drop so 'now - drop_next'
           // is a good approximation of the time from the last drop
           // until now.) Implementations vary slightly here; this is
           // the Linux version, which is more widely deployed and
           // tested.



Or summarized: this issue is not addressed in text, which instead just ment=
ions that the issue exists, there are several versions of this logic and he=
re is one of them.


As it reads now, this "reentering" should be a tiny insignificant part of t=
he algorithm that (1) doesn't happen in "most" of the cases, and (2) when i=
t does happen, it will only "slightly" affect "performance" (with proper de=
finitions of "most", "slightly", and "performance").


Is it really so?


Based on our evaluations, with pure CoDel (without FQ-CoDel), "reentering" =
is actually a common case. I think Dave and Toke should have more experimen=
tal results to answer this question. (I included Dave in CC)


If it is so, I still think that the "common case" should be at the very lea=
st explained in text too.


Issue 2: The "initial drop spacing" is inconsistent


Regarding the "initial drop spacing" there are following lines in the docum=
ent.


between pages 11 and 12:

>and so the initial drop spacing
>SHOULD be set to the estimator's interval plus twice the target
>(i.e., initial drop spacing =3D 1.1 * interval)

page 12:

>the initial next
>drop spacing is intended to be long enough to give the endpoints time
>to react to the single drop so SHOULD be set to a value of 1.1 times
>the interval

page 16:


>and the initial drop
>spacing is also set to interval.

and it is also set to one interval in pseudo-code, which I believe is this =
line:


> drop_next_ =3D control_law(now, count_);


Note: I think there was an email on CoDel mailing list (https://lists.buffe=
rbloat.net/pipermail/codel/) about this issue.



________________________________
From: Andrew Mcgregor <andrewmcgr@google.com>
Sent: Friday, June 3, 2016 3:50 AM
To: Goltsman, Polina
Cc: Wesley Eddy; aqm@ietf.org; Jana Iyengar
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt

Thank you Polina.

We have incorporated most of these (sometimes with slightly different solut=
ions than you suggested), and a new draft will be out very shortly.  I felt=
 that by far the most important was the different way the Linux implementat=
ion addressed re-entry to the dropping state, and the pseudocode now does e=
xactly that.

Andrew

On 22 March 2016 at 08:10, Polina Goltsman <polina.goltsman@student.kit.edu=
<mailto:polina.goltsman@student.kit.edu>> wrote:
Dear Wesley, Dear All,

First of all our feedback regarding different "re-entering dropping state" =
in the document and in the Linux implementation (http://www.ietf.org/mail-a=
rchive/web/aqm/current/msg01686.html) was not addressed.

As FQ-CoDel  relies on CoDel, this issue is also (partly) relevant for the =
FQ-CoDel document. In the introduction FQ-CoDel references ns-3 and Linux i=
mplementations where the first one uses the re-entering logic from the CoDe=
l document while the second from CoDel Linux implementation. The algorithm =
that has seen widespread testing according to Section 7 is (I suppose) the =
Linux version. Is this situation acceptable for an algorithm specification?

[since this comment was supposed to be sent before the end of 2015, feel fr=
ee to (silently) ignore it]
Second, unlike Rasool Al-Saadi (see http://www.ietf.org/mail-archive/web/aq=
m/current/msg01693.html) I do not like the document. Although I agree that =
the pseudocode is sufficient to create a working implementation, however, i=
n my opinion, the rest of the document makes implementing CoDeL more confus=
ing (at least without reading [CODEL2012] first). Is it normal for a RFC, w=
hich, as I assume, should primarily contain an algorithm specification to c=
ontain the algorithm specification ONLY in form of pseudocode?
These are two items that I found the most confusing:
(1) section 3 introduces estimator, setpoint, and control loop which are no=
t clearly distinguishable in pseudocode. It would be nice if section 4 expl=
ained how the three entities transition into the routines in pseudocode.
(2) IMHO, there is some missing explanations. For example, the document nev=
er says how exactly "bad"/persistent queue is determined. The document says=
 in Section 3.1 that the minimum is tracked, but it never says that persist=
ent queue is when the minimum is above setpoint/target for at least an inte=
rval. As another example, Section 3.3 could say how the controller looks li=
ke exactly.

I also have some small comments and nits regarding the latest version (-03)=
:

---

in the whole document there are several places of using "we" to (supposedly=
) refer to the authors. According to Bob Briscoe's message (http://www.ietf=
.org/mail-archive/web/aqm/current/msg01805.html), "we" do not necessary ref=
er to the authors in a rfc. Are these "we"s ok?

---

abstract:

This document describes a general framework called CoDel

IMHO the word "framework" is too overloaded and too broad. It might be bett=
er to call CoDel an AQM or something that has "queue" in it. Other ways (e.=
g., http://queue.acm.org/detail.cfm?id=3D2839461) probably also use it to c=
ontrol a queue.

---

section 1:

(nit)
   Despite this
   awareness, the problem has only gotten worse as Moore's Law growth in
   memory density fueled an exponential increase in buffer pool size.
   Efforts to deploy AQM have been frustrated by difficult configuration
   and negative impact on network utilization.  This problem, recently
   christened "bufferbloat", [TSV2011] [BB2011] has become increasingly
   important throughout the Internet but particularly at the consumer
   edge.

it reads like "this problem" refers to "efforts to deploy AQM" and not to "=
increase in buffer pool size"

---

(nit)

   To understand
   queue management, it is critical to understand the difference between
   the necessary, useful "good" queue, and the counterproductive "bad"
   queue.

strictly speaking "good" and "bad" queue were not defined either before or =
in this sentence (it is explained in section 3 or in [CODEL2012])

---

(nit)

   o  treat "good queue" and "bad queue" differently, that is, keep
      delay low while permitting necessary bursts of traffic

(1) see above
(2) I would say that the goal is "to keep delay low while ..." and to treat=
 good/bad queue differently is a mechanism used to achieve the goal

---

section 2:

(nit)
      In this document, the characters ">>" preceding an indented line(s)
   indicates a compliance requirement statement using the key words
   listed above.  This convention aids reviewers in quickly identifying
   or finding the explicit compliance requirements of this RFC.

I maybe missing something, but I can't find any use of "characters ">>"" an=
ywhere except this line in any (txt,html, pdf) of the versions.

---

section 3:

generally IMHO it would be nice to introduce interval and target so it is m=
ore obvious that these are parameters (or some other kind of variables). In=
terval is more or less obvious (although imho should be more explicit). The=
 target is first introduced at the end of section 3.2 as:


   This results in a particularly simple form for the
   setpoint: the ideal range for the permitted standing queue is between
   5% and 10% of the TCP connection's RTT.  Thus target is simply 5% of
   the interval of section 3.1<https://tools.ietf.org/html/draft-ietf-aqm-c=
odel-03#section-3.1>.


It was not previously defined that target is a parameter with a meaning of =
setpoint. I don't think that this is clear for people who do not know what =
target is in advance.

[It seems that it was intended to introduce interval and target parameters =
in section 4]

---


   Instead of averages
   we recommend tracking the minimum sojourn time, then,

"we recommend" or "CoDeL does"? According to the last paragraph in Section =
3.1, tracking minimum is a core part of CoDeL:

   These two innovations, use of sojourn time as observed values and the
   local minimum as the statistic to monitor queue congestion are key to
   CoDel's estimator building block.

---

   Since the peak queue delay is simply f r, power is solely a function

is it f/r ?

---
(nit)

   The only remaining building block needed for a basic AQM is a
   'control loop' algorithm to effectively drive the queueing system
   from any 'persistent queue above target' state to a state where the
   persistent queue is below target.

It would be nice to explicitly state that the "persistent queue above targe=
t" is indication of "bad" queue/congestion somewhere above or in this line.=
 I think that these terms are used in [codel2012] but were not used before =
in the document. See also the next item.

---
In section 3.3 there are three "states":

  *

'persistent queue above target' state

  *

'has persistent queue' state

  *

 dropping state

It would be nice to a) use the same one b) explain how CoDeL gets in/out of=
 there

---

(probably nit)

   When
   the minimum sojourn time first crosses the target and CoDel drops a
   packet, the earliest the controller could see the effect of the drop

from current sentence it could be inferred that the two events occur at the=
 same time which is not the case. Maybe " when persistent queue above targe=
t is detected" would be better.

---

   variation is less than the target, and so the initial drop spacing
   SHOULD be set to the estimator's interval plus twice the target
   (i.e., initial drop spacing =3D 1.1 * interval) to ensure that the

this contradicts the pseudocode:

   dodequeue_result codel_queue_t::dodequeue(time_t now)

        ...
        first_above_time_ =3D now + interval_;

section 4:


   To keep from
   making drops when it would starve the output link, CoDel makes
   another check before dropping to see if at least an MTU worth of
   bytes remains in the buffer. If not, the packet SHOULD NOT be
   dropped and, currently, CoDel exits the drop state.

what is the meaning of "currently" ? in the current experimental version?

---
(nit)

   In practice, this value is not known or measured (though see
   Section 6.2<https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-=
6.2> for an application where interval is measured.

missing closing bracket

Section 5 / pseudocode: (all nits)

The variables are introduced as first_time_above, dropping, but are used as=
 first_time_above_, dropping_ (with underscores at end).

---

Packet* CoDelQueue::dequeue()


a pointer to current packet was introduced as packet_t

---

double now =3D clock();;


time units were supposed to be of type time_t

---

dodequeueResult r =3D dodequeue(now);

compare with: typedef struct { ... } dodequeue_result below (present in two=
 routines)

---

dodequeueResult r =3D { NULL, queue_t::dequeue() };

shouldn't it be {queue_t::dequeue(), false} ?


Best Regards,
Polina


* note the reply by Dave Taht: http://www.ietf.org/mail-archive/web/aqm/cur=
rent/msg01687.html

On 03/20/2016 04:35 PM, Wesley Eddy wrote:
It looks like the WGLC feedback on the document body is incorporated, so th=
is is good.

Is there a reason to stay with Informational and not Experimental like we'v=
e done with PIE an d FQ-CoDel?

Also, idnits has some problems with the references that should be fixed (e.=
g. "NATAL2010" is probably supposed to be "NETAL2010"):
https://tools.ietf.org/idnits?url=3Dhttps://tools.ietf.org/id/draft-ietf-aq=
m-codel-03.txt


_______________________________________________
aqm mailing list
aqm@ietf.org<mailto:aqm@ietf.org>
https://www.ietf.org/mailman/listinfo/aqm




--
Andrew McGregor | SRE | andrewmcgr@google.com<mailto:andrewmcgr@google.com>=
 | +61 4 1071 2221

--_000_146789204844564588studentkitedu_
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Diso-8859-=
1">
<style type=3D"text/css" style=3D"display:none"><!--P{margin-top:0;margin-b=
ottom:0;} --></style>
</head>
<body dir=3D"ltr" style=3D"font-size:12pt;color:#000000;background-color:#F=
FFFFF;font-family:Calibri,Arial,Helvetica,sans-serif;">
<p>Dear Andrew, Dear Wesley,</p>
<p><br>
</p>
<p>First of all, sorry for the long waiting time. I still haven't processed=
 &quot;nits&quot; but I have 2 major issues.</p>
<p><br>
</p>
<p><strong>Issue 1: </strong>This issue is still related to the reentering =
dropping state.</p>
<p><br>
</p>
<p>As of the latest I-D, the &quot;reentering dropping state&quot; &nbsp;ap=
pears in form of one sentence in the text and a small code block in the pse=
udo-code.</p>
<p><br>
</p>
<p>The text is on page 13:</p>
<pre class=3D"newpage">&gt;Additional logic prevents re-=0A=
&gt;entering the dropping state too soon after exiting it and resumes the=
=0A=
&gt;dropping state at a recent control level, if one exists.</pre>
<p>The comment to the pseudo-code on page 20:</p>
<p><br>
</p>
<pre class=3D"newpage">           // If min went above target close to when=
 it last went=0A=
           // below, assume that the drop rate that controlled the=0A=
           // queue on the last cycle is a good starting point to=0A=
           // control it now. ('drop_next' will be at most 'interval'=0A=
           // later than the time of the last drop so 'now - drop_next'=0A=
           // is a good approximation of the time from the last drop=0A=
           // until now.) <strong>Implementations vary slightly here; this =
is=0A=
           // the Linux version, which is more widely deployed and=0A=
           // tested.<br><br></strong></pre>
<p><br>
</p>
<p>Or summarized: this issue is not addressed in text, which instead just m=
entions that the issue exists, there are several versions of this logic and=
 here is one of them.</p>
<p><br>
</p>
<p>As it reads now, this &quot;reentering&quot; should be a tiny insignific=
ant part of the algorithm that (1) doesn't happen in &quot;most&quot; of th=
e cases, and (2) when it does happen, it will only &quot;slightly&quot; aff=
ect &quot;performance&quot; (with proper&nbsp;definitions of &quot;most&quo=
t;, &quot;slightly&quot;,
 and &quot;performance&quot;).</p>
<p><br>
</p>
<p>Is it really&nbsp;so?&nbsp;</p>
<p><br>
</p>
<p>Based on our evaluations, with pure CoDel (without FQ-CoDel), &quot;reen=
tering&quot;&nbsp;is actually a common case. I think Dave and Toke should h=
ave more&nbsp;experimental results to&nbsp;answer this question. (I include=
d Dave in CC)<br>
</p>
<p><br>
</p>
<p>If it is so, I&nbsp;still think that&nbsp;the &quot;common case&quot; sh=
ould be at the very least explained in text too.
</p>
<p><br>
</p>
<p><strong>Issue 2:</strong> The &quot;initial drop spacing&quot; is incons=
istent<br>
</p>
<p><br>
</p>
<p>Regarding the &quot;initial drop spacing&quot; there are following lines=
 in the document.</p>
<p><br>
</p>
<p>between pages 11 and 12:</p>
<pre class=3D"newpage">&gt;and so the initial drop spacing <br>&gt;SHOULD b=
e set to the estimator's <strong>interval plus twice the target</strong>=0A=
&gt;(i.e., initial drop spacing =3D 1.1 * interval)<br><br>page 12:<br><pre=
 class=3D"newpage">&gt;the initial next=0A=
&gt;drop spacing is intended to be long enough to give the endpoints time=
=0A=
&gt;to react to the single drop so SHOULD be set to a value <strong>of 1.1 =
times=0A=
&gt;the interval</strong></pre></pre>
<p>page 16:</p>
<p><br>
</p>
<pre class=3D"newpage">&gt;and the initial drop=0A=
&gt;spacing is also set to <strong>interval</strong>.</pre>
<p>and it is also set to<strong> one interval</strong> in pseudo-code, whic=
h I believe is this line:</p>
<p><br>
</p>
<p>&gt; drop_next_ =3D control_law(now, count_);</p>
<p><br>
</p>
<p>Note: I think there was an email on CoDel mailing list (<a href=3D"https=
://lists.bufferbloat.net/pipermail/codel/">https://lists.bufferbloat.net/pi=
permail/codel/</a>) about this issue.
<br>
</p>
<p><br>
</p>
<p><br>
</p>
<div style=3D"color: rgb(33, 33, 33);">
<hr tabindex=3D"-1" style=3D"display:inline-block; width:98%">
<div id=3D"divRplyFwdMsg" dir=3D"ltr"><font style=3D"font-size:11pt" face=
=3D"Calibri, sans-serif" color=3D"#000000"><b>From:</b> Andrew Mcgregor &lt=
;andrewmcgr@google.com&gt;<br>
<b>Sent:</b> Friday, June 3, 2016 3:50 AM<br>
<b>To:</b> Goltsman, Polina<br>
<b>Cc:</b> Wesley Eddy; aqm@ietf.org; Jana Iyengar<br>
<b>Subject:</b> Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt</font>
<div>&nbsp;</div>
</div>
<div>
<div dir=3D"ltr">Thank you Polina.
<div><br>
</div>
<div>We have incorporated most of these (sometimes with slightly different =
solutions than you suggested), and a new draft will be out very shortly.&nb=
sp; I felt that by far the most important was the different way the Linux i=
mplementation addressed re-entry to the
 dropping state, and the pseudocode now does exactly that.</div>
<div><br>
</div>
<div>Andrew</div>
</div>
<div class=3D"gmail_extra"><br>
<div class=3D"gmail_quote">On 22 March 2016 at 08:10, Polina Goltsman <span=
 dir=3D"ltr">
&lt;<a href=3D"mailto:polina.goltsman@student.kit.edu" target=3D"_blank">po=
lina.goltsman@student.kit.edu</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex; border-left:1=
px #ccc solid; padding-left:1ex">
<div bgcolor=3D"#FFFFFF">Dear Wesley, Dear All,<br>
<br>
First of all our feedback regarding different &quot;re-entering dropping st=
ate&quot; in the document and in the Linux implementation (<a href=3D"http:=
//www.ietf.org/mail-archive/web/aqm/current/msg01686.html" target=3D"_blank=
">http://www.ietf.org/mail-archive/web/aqm/current/msg01686.html</a>)
 was not addressed. <br>
<br>
As FQ-CoDel&nbsp; relies on CoDel, this issue is also (partly) relevant for=
 the FQ-CoDel document. In the introduction FQ-CoDel references ns-3 and Li=
nux implementations where the first one uses the re-entering logic from the=
 CoDel document while the second from
 CoDel Linux implementation. The algorithm that has seen widespread testing=
 according to Section 7 is (I suppose) the Linux version. Is this situation=
 acceptable for an algorithm specification?
<br>
<br>
<i>[since this comment was supposed to be sent before the end of 2015, feel=
 free to (silently) ignore it]</i>
<br>
Second, unlike Rasool Al-Saadi (see <a href=3D"http://www.ietf.org/mail-arc=
hive/web/aqm/current/msg01693.html" target=3D"_blank">
http://www.ietf.org/mail-archive/web/aqm/current/msg01693.html</a>) I do no=
t like the document. Although I agree that the pseudocode is sufficient to =
create a working implementation, however, in my opinion, the rest of the do=
cument makes implementing CoDeL
 more confusing (at least without reading [CODEL2012] first). Is it normal =
for a RFC, which, as I assume, should primarily contain an algorithm specif=
ication to contain the algorithm specification ONLY in form of pseudocode? =
&nbsp;
<br>
These are two items that I found the most confusing:<br>
(1) section 3 introduces estimator, setpoint, and control loop which are no=
t clearly distinguishable in pseudocode. It would be nice if section 4 expl=
ained how the three entities transition into the routines in pseudocode.
<br>
(2) IMHO, there is some missing explanations. For example, the document nev=
er says how exactly &quot;bad&quot;/persistent queue is determined. The doc=
ument says in Section 3.1 that the minimum is tracked, but it never says th=
at persistent queue is when the minimum is
 above setpoint/target for at least an interval. As another example, Sectio=
n 3.3 could say how the controller looks like exactly.<br>
<br>
I also have some small comments and nits regarding the latest version (-03)=
:<br>
<br>
---<br>
<br>
in the whole document there are several places of using &quot;we&quot; to (=
supposedly) refer to the authors. According to Bob Briscoe's message (<a hr=
ef=3D"http://www.ietf.org/mail-archive/web/aqm/current/msg01805.html" targe=
t=3D"_blank">http://www.ietf.org/mail-archive/web/aqm/current/msg01805.html=
</a>),
 &quot;we&quot; do not necessary refer to the authors in a rfc. Are these &=
quot;we&quot;s ok?<br>
<br>
---<br>
<br>
abstract:<span class=3D""><br>
<pre>This document describes a general <b>framework</b> called CoDel </pre>
</span>IMHO the word &quot;framework&quot; is too overloaded and too broad.=
 It might be better to call CoDel an AQM or something that has &quot;queue&=
quot; in it. Other ways (e.g.,
<a href=3D"http://queue.acm.org/detail.cfm?id=3D2839461" target=3D"_blank">=
http://queue.acm.org/detail.cfm?id=3D2839461</a>) probably also use it to c=
ontrol a queue.<br>
<br>
---<br>
<br>
section 1:<br>
<br>
(nit)<br>
<tt>&nbsp;&nbsp; Despite this</tt><tt><br>
</tt><tt>&nbsp;&nbsp; awareness, the problem has only gotten worse as Moore=
's Law growth in</tt><tt><br>
</tt><tt>&nbsp;&nbsp; memory density fueled an exponential increase in buff=
er pool size.</tt><tt><br>
</tt><tt>&nbsp;&nbsp; Efforts to deploy AQM have been frustrated by difficu=
lt configuration</tt><tt><br>
</tt><tt>&nbsp;&nbsp; and negative impact on network utilization.&nbsp; <b>=
This problem,</b> recently</tt><tt><br>
</tt><tt>&nbsp;&nbsp; christened &quot;bufferbloat&quot;, [TSV2011] [BB2011=
] has become increasingly</tt><tt><br>
</tt><tt>&nbsp;&nbsp; important throughout the Internet but particularly at=
 the consumer</tt><tt><br>
</tt><tt>&nbsp;&nbsp; edge. </tt><br>
<br>
it reads like &quot;this problem&quot; refers to &quot;efforts to deploy AQ=
M&quot; and not to &quot;increase in buffer pool size&quot;<br>
<br>
---<br>
<br>
(nit) <br>
<pre>   To understand=0A=
   queue management, it is critical to understand the difference between=0A=
   the necessary, useful <b>&quot;good&quot; queue</b>, and the counterprod=
uctive <b>&quot;bad&quot;=0A=
   queue</b>.</pre>
strictly speaking &quot;good&quot; and &quot;bad&quot; queue were not defin=
ed either before or in this sentence (it is explained in section 3 or in [C=
ODEL2012])<br>
<br>
---<br>
<br>
(nit)<br>
<pre>   o  treat &quot;good queue&quot; and &quot;bad queue&quot; different=
ly, that is, keep=0A=
      delay low while permitting necessary bursts of traffic</pre>
(1) see above<br>
(2) I would say that the goal is &quot;to keep delay low while ...&quot; an=
d to treat good/bad queue differently is a mechanism used to achieve the go=
al<br>
<br>
---<br>
<br>
section 2:<br>
<br>
(nit)<tt><br>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; In this document, <b>the characters &quot;&g=
t;&gt;&quot; preceding an indented line(s)</b></tt><tt><br>
</tt><tt>&nbsp;&nbsp; indicates a compliance requirement statement using th=
e key words</tt><tt><br>
</tt><tt>&nbsp;&nbsp; listed above.&nbsp; This convention aids reviewers in=
 quickly identifying</tt><tt><br>
</tt><tt>&nbsp;&nbsp; or finding the explicit compliance requirements of th=
is RFC.</tt><br>
<br>
I maybe missing something, but I can't find any use of &quot;characters &qu=
ot;&gt;&gt;&quot;&quot; anywhere except this line in any (txt,html, pdf) of=
 the versions.<br>
<br>
---<br>
<br>
section 3: <br>
<br>
generally IMHO it would be nice to introduce <i>interval</i> and <i>target<=
/i> so it is more obvious that these are parameters (or some other kind of =
variables). Interval is more or less obvious (although imho should be more =
explicit). The target is first introduced
 at the end of section 3.2 as:<br>
<br>
<pre>   This results in a particularly simple form for the=0A=
   setpoint: the ideal range for the permitted standing queue is between=0A=
   5% and 10% of the TCP connection's RTT.  Thus <b>target</b> is simply 5%=
 of=0A=
   the interval of <a href=3D"https://tools.ietf.org/html/draft-ietf-aqm-co=
del-03#section-3.1" target=3D"_blank">section 3.1</a>.=0A=
</pre>
It was not previously defined that <i>target</i> is a parameter with a mean=
ing of
<i>setpoint</i>. I don't think that this is clear for people who do not kno=
w what target is in advance.<br>
<br>
[It seems that it was intended to introduce <i>interval</i> and <i>target</=
i> parameters in section 4]<br>
<br>
---<br>
<br>
<pre>   Instead of averages=0A=
   <b>we recommend</b> tracking the minimum sojourn time, then, </pre>
&quot;we recommend&quot; or &quot;CoDeL does&quot;? According to the last p=
aragraph in Section 3.1, tracking minimum is a core part of CoDeL:<br>
<pre>   These two innovations, use of sojourn time as observed values <b>an=
d the=0A=
   local minimum as the statistic to monitor queue congestion</b> are key t=
o=0A=
   CoDel's estimator building block.</pre>
---<br>
<pre>   Since the peak queue delay<b> is simply f r</b>, power is solely a =
function</pre>
is it f/r ?<br>
<br>
---<br>
(nit)<br>
<pre>   The only remaining building block needed for a basic AQM is a=0A=
   'control loop' algorithm to effectively drive the queueing system=0A=
   from any <b>'persistent queue above target'</b> state to a state where t=
he=0A=
   <b>persistent queue is below target</b>.</pre>
It would be nice to explicitly state that the &quot;persistent queue above =
target&quot; is indication of &quot;bad&quot; queue/congestion somewhere ab=
ove or in this line. I think that these terms are used in [codel2012] but w=
ere not used before in the document. See also the next
 item.<br>
<br>
---<br>
In section 3.3 there are three &quot;states&quot;:<br>
<ul>
<li>
<pre>'persistent queue above target' state</pre>
</li><li>
<pre>'has persistent queue' state</pre>
</li><li>
<pre> dropping state</pre>
</li></ul>
<p>It would be nice to a) use the same one b) explain how CoDeL gets in/out=
 of there<br>
</p>
<p>---<br>
</p>
(probably nit)<br>
<pre>   When=0A=
   <b>the minimum sojourn time first crosses the target</b> and <b>CoDel dr=
ops a=0A=
   packet</b>, the earliest the controller could see the effect of the drop=
</pre>
from current sentence it could be inferred that the two events occur at the=
 same time which is not the case. Maybe &quot; when persistent queue above =
target is detected&quot; would be better.<br>
<br>
---<br>
<pre>   variation is less than the target, and so the <b>initial drop spaci=
ng=0A=
   SHOULD be set to the estimator's interval plus twice the target=0A=
   (i.e., initial drop spacing =3D 1.1 * interval)</b> to ensure that the</=
pre>
this contradicts the pseudocode:<br>
<pre>   dodequeue_result codel_queue_t::dodequeue(time_t now)=0A=
=0A=
	...=0A=
	first_above_time_ =3D now &#43; interval_;</pre>
section 4:<br>
<br>
<pre>   To keep from=0A=
   making drops when it would starve the output link, CoDel makes=0A=
   another check before dropping to see if at least an MTU worth of=0A=
   bytes remains in the buffer. If not, the packet SHOULD NOT be=0A=
   dropped and, <b>currently</b>, CoDel exits the drop state. </pre>
what is the meaning of &quot;currently&quot; ? in the current experimental =
version?<br>
<br>
---<br>
(nit)<br>
<pre>   In practice, this value is not known or measured <b>(</b>though see=
=0A=
   <a href=3D"https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-6=
.2" target=3D"_blank">Section 6.2</a> for an application where interval is =
measured.</pre>
missing closing bracket<br>
<br>
Section 5 / pseudocode: (all nits)<br>
<br>
The variables are introduced as first_time_above, dropping, but are used as=
 first_time_above_, dropping_ (with underscores at end).&nbsp;
<br>
<br>
---<br>
<pre><b>Packet</b>* CoDelQueue::dequeue()=0A=
</pre>
a pointer to current packet was introduced as packet_t<br>
<pre><b>---=0A=
=0A=
double</b> now =3D clock();;=0A=
</pre>
time units were supposed to be of type time_t<br>
<pre><b>---=0A=
=0A=
dodequeueResult</b> r =3D dodequeue(now);</pre>
compare with: typedef struct { ... } <b>dodequeue_result</b> below (present=
 in two routines)<br>
<pre>---=0A=
=0A=
dodequeueResult r =3D {<b> NULL, queue_t::dequeue() </b>};</pre>
shouldn't it be {queue_t::dequeue(), false} ?<br>
<br>
<br>
Best Regards,<br>
Polina<br>
<br>
<br>
* note the reply by Dave Taht: <a href=3D"http://www.ietf.org/mail-archive/=
web/aqm/current/msg01687.html" target=3D"_blank">
http://www.ietf.org/mail-archive/web/aqm/current/msg01687.html</a><span cla=
ss=3D""><br>
<br>
<div>On 03/20/2016 04:35 PM, Wesley Eddy wrote:<br>
</div>
<blockquote type=3D"cite">It looks like the WGLC feedback on the document b=
ody is incorporated, so this is good.
<br>
<br>
Is there a reason to stay with Informational and not Experimental like we'v=
e done with PIE an d FQ-CoDel?
<br>
<br>
Also, idnits has some problems with the references that should be fixed (e.=
g. &quot;NATAL2010&quot; is probably supposed to be &quot;NETAL2010&quot;):
<br>
<a href=3D"https://tools.ietf.org/idnits?url=3Dhttps://tools.ietf.org/id/dr=
aft-ietf-aqm-codel-03.txt" target=3D"_blank">https://tools.ietf.org/idnits?=
url=3Dhttps://tools.ietf.org/id/draft-ietf-aqm-codel-03.txt</a></blockquote=
>
<br>
</span></div>
<br>
_______________________________________________<br>
aqm mailing list<br>
<a href=3D"mailto:aqm@ietf.org">aqm@ietf.org</a><br>
<a href=3D"https://www.ietf.org/mailman/listinfo/aqm" rel=3D"noreferrer" ta=
rget=3D"_blank">https://www.ietf.org/mailman/listinfo/aqm</a><br>
<br>
</blockquote>
</div>
<br>
<br clear=3D"all">
<div><br>
</div>
-- <br>
<div class=3D"gmail_signature">
<div dir=3D"ltr"><span style=3D"color:rgb(85,85,85); font-family:sans-serif=
; font-size:small; line-height:1.5em; border-width:2px 0px 0px; border-styl=
e:solid; border-color:rgb(213,15,37); padding-top:2px; margin-top:2px">Andr=
ew McGregor&nbsp;|</span><span style=3D"color:rgb(85,85,85); font-family:sa=
ns-serif; font-size:small; line-height:1.5em; border-width:2px 0px 0px; bor=
der-style:solid; border-color:rgb(51,105,232); padding-top:2px; margin-top:=
2px">&nbsp;SRE&nbsp;|</span><span style=3D"color:rgb(85,85,85); font-family=
:sans-serif; font-size:small; line-height:1.5em; border-width:2px 0px 0px; =
border-style:solid; border-color:rgb(0,153,57); padding-top:2px; margin-top=
:2px">&nbsp;<a href=3D"mailto:andrewmcgr@google.com" target=3D"_blank">andr=
ewmcgr@google.com</a>&nbsp;|</span><span style=3D"color:rgb(85,85,85); font=
-family:sans-serif; font-size:small; line-height:1.5em; border-width:2px 0p=
x 0px; border-style:solid; border-color:rgb(238,178,17); padding-top:2px; m=
argin-top:2px">&nbsp;&#43;61
 4 1071 2221</span><br>
</div>
</div>
</div>
</div>
</div>
</body>
</html>

--_000_146789204844564588studentkitedu_--


From nobody Thu Jul  7 05:34:21 2016
Return-Path: <menth@uni-tuebingen.de>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C18D912D752 for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 05:34:19 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -4.026
X-Spam-Level: 
X-Spam-Status: No, score=-4.026 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_LOW=-0.7, RP_MATCHES_RCVD=-1.426] autolearn=ham autolearn_force=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 rbboItL-4vxn for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 05:34:15 -0700 (PDT)
Received: from mx03.uni-tuebingen.de (mx03.uni-tuebingen.de [134.2.5.213]) (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 1BD1812D688 for <aqm@ietf.org>; Thu,  7 Jul 2016 05:34:14 -0700 (PDT)
Received: from [134.2.11.131] (chaos.informatik.uni-tuebingen.de [134.2.11.131]) by mx03.uni-tuebingen.de (Postfix) with ESMTPSA id A988D782A7; Thu,  7 Jul 2016 14:34:12 +0200 (CEST)
To: "Goltsman, Polina" <polina.goltsman@student.kit.edu>, Andrew Mcgregor <andrewmcgr@google.com>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu>
From: Michael Menth <menth@uni-tuebingen.de>
Message-ID: <577E4C53.9030909@uni-tuebingen.de>
Date: Thu, 7 Jul 2016 14:34:27 +0200
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.7.2
MIME-Version: 1.0
In-Reply-To: <1467892048445.64588@student.kit.edu>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 8bit
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/x-9I0htiR7YetDe4fbsjJDbLnRo>
Cc: Fabian Schwarzkopf <f.schwarzkopf@yahoo.de>, Wesley Eddy <wes@mti-systems.com>, Dave Taht <dave.taht@gmail.com>, Jana Iyengar <jri@google.com>, "Bless, Roland \(TM\)" <roland.bless@kit.edu>, "aqm@ietf.org" <aqm@ietf.org>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 12:34:20 -0000

Dear Polina, dear all,

Am 07.07.2016 um 13:49 schrieb Goltsman, Polina:
> Dear Andrew, Dear Wesley,
> 
> 
> First of all, sorry for the long waiting time. I still haven't processed
> "nits" but I have 2 major issues.
> 
> 
> *Issue 1: *This issue is still related to the reentering dropping state.
> 
> 
> As of the latest I-D, the "reentering dropping state"  appears in form
> of one sentence in the text and a small code block in the pseudo-code.
> 
> 
> The text is on page 13:
> 
>>Additional logic prevents re-
>>entering the dropping state too soon after exiting it and resumes the
>>dropping state at a recent control level, if one exists.
> 
> The comment to the pseudo-code on page 20:
> 
> 
>            // If min went above target close to when it last went
>            // below, assume that the drop rate that controlled the
>            // queue on the last cycle is a good starting point to
>            // control it now. ('drop_next' will be at most 'interval'
>            // later than the time of the last drop so 'now - drop_next'
>            // is a good approximation of the time from the last drop
>            // until now.) *Implementations vary slightly here; this is // the Linux version, which
> is more widely deployed and // tested.
> 
> *
> 
> 
> Or summarized: this issue is not addressed in text, which instead just
> mentions that the issue exists, there are several versions of this logic
> and here is one of them.
> 
> 
> As it reads now, this "reentering" should be a tiny insignificant part
> of the algorithm that (1) doesn't happen in "most" of the cases, and (2)
> when it does happen, it will only "slightly" affect "performance" (with
> proper definitions of "most", "slightly", and "performance").
> 
> 
> Is it really so? 
> 
> 
> Based on our evaluations, with pure CoDel (without FQ-CoDel),
> "reentering" is actually a common case. I think Dave and Toke should
> have more experimental results to answer this question. (I included Dave
> in CC)


We also studied this issue about one or two years ago. The algorithm
controlling the reentering of the dropping mode may have a significant
impact on control variables, drop rates, drop pattern depending on the
traffic. It can also impact utilization. Most interesting is probably
Fig. 6 on page 7 in
https://atlas.informatik.uni-tuebingen.de/~menth/papers/Menth16e.pdf

Kind regards,

Michael



> 
> 
> If it is so, I still think that the "common case" should be at the very
> least explained in text too.
> 
> 
> *Issue 2:* The "initial drop spacing" is inconsistent
> 
> 
> Regarding the "initial drop spacing" there are following lines in the
> document.
> 
> 
> between pages 11 and 12:
> 
>>and so the initial drop spacing 
>>SHOULD be set to the estimator's *interval plus twice the target*
>>(i.e., initial drop spacing = 1.1 * interval)
> 
> page 12:
> 
>>the initial next
>>drop spacing is intended to be long enough to give the endpoints time
>>to react to the single drop so SHOULD be set to a value *of 1.1 times >the interval*
> 
> page 16:
> 
> 
>>and the initial drop
>>spacing is also set to *interval*.
> 
> and it is also set to*one interval* in pseudo-code, which I believe is
> this line:
> 
> 
>> drop_next_ = control_law(now, count_);
> 
> 
> Note: I think there was an email on CoDel mailing list
> (https://lists.bufferbloat.net/pipermail/codel/) about this issue.
> 
> 
> 
> ------------------------------------------------------------------------
> *From:* Andrew Mcgregor <andrewmcgr@google.com>
> *Sent:* Friday, June 3, 2016 3:50 AM
> *To:* Goltsman, Polina
> *Cc:* Wesley Eddy; aqm@ietf.org; Jana Iyengar
> *Subject:* Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
>  
> Thank you Polina.
> 
> We have incorporated most of these (sometimes with slightly different
> solutions than you suggested), and a new draft will be out very
> shortly.  I felt that by far the most important was the different way
> the Linux implementation addressed re-entry to the dropping state, and
> the pseudocode now does exactly that.
> 
> Andrew
> 
> On 22 March 2016 at 08:10, Polina Goltsman
> <polina.goltsman@student.kit.edu
> <mailto:polina.goltsman@student.kit.edu>> wrote:
> 
>     Dear Wesley, Dear All,
> 
>     First of all our feedback regarding different "re-entering dropping
>     state" in the document and in the Linux implementation
>     (http://www.ietf.org/mail-archive/web/aqm/current/msg01686.html) was
>     not addressed.
> 
>     As FQ-CoDel  relies on CoDel, this issue is also (partly) relevant
>     for the FQ-CoDel document. In the introduction FQ-CoDel references
>     ns-3 and Linux implementations where the first one uses the
>     re-entering logic from the CoDel document while the second from
>     CoDel Linux implementation. The algorithm that has seen widespread
>     testing according to Section 7 is (I suppose) the Linux version. Is
>     this situation acceptable for an algorithm specification?
> 
>     /[since this comment was supposed to be sent before the end of 2015,
>     feel free to (silently) ignore it]/
>     Second, unlike Rasool Al-Saadi (see
>     http://www.ietf.org/mail-archive/web/aqm/current/msg01693.html) I do
>     not like the document. Although I agree that the pseudocode is
>     sufficient to create a working implementation, however, in my
>     opinion, the rest of the document makes implementing CoDeL more
>     confusing (at least without reading [CODEL2012] first). Is it normal
>     for a RFC, which, as I assume, should primarily contain an algorithm
>     specification to contain the algorithm specification ONLY in form of
>     pseudocode?  
>     These are two items that I found the most confusing:
>     (1) section 3 introduces estimator, setpoint, and control loop which
>     are not clearly distinguishable in pseudocode. It would be nice if
>     section 4 explained how the three entities transition into the
>     routines in pseudocode.
>     (2) IMHO, there is some missing explanations. For example, the
>     document never says how exactly "bad"/persistent queue is
>     determined. The document says in Section 3.1 that the minimum is
>     tracked, but it never says that persistent queue is when the minimum
>     is above setpoint/target for at least an interval. As another
>     example, Section 3.3 could say how the controller looks like exactly.
> 
>     I also have some small comments and nits regarding the latest
>     version (-03):
> 
>     ---
> 
>     in the whole document there are several places of using "we" to
>     (supposedly) refer to the authors. According to Bob Briscoe's
>     message
>     (http://www.ietf.org/mail-archive/web/aqm/current/msg01805.html),
>     "we" do not necessary refer to the authors in a rfc. Are these "we"s ok?
> 
>     ---
> 
>     abstract:
> 
>     This document describes a general *framework* called CoDel 
> 
>     IMHO the word "framework" is too overloaded and too broad. It might
>     be better to call CoDel an AQM or something that has "queue" in it.
>     Other ways (e.g., http://queue.acm.org/detail.cfm?id=2839461)
>     probably also use it to control a queue.
> 
>     ---
> 
>     section 1:
> 
>     (nit)
>        Despite this
>        awareness, the problem has only gotten worse as Moore's Law growth in
>        memory density fueled an exponential increase in buffer pool size.
>        Efforts to deploy AQM have been frustrated by difficult configuration
>        and negative impact on network utilization.  *This problem,* recently
>        christened "bufferbloat", [TSV2011] [BB2011] has become increasingly
>        important throughout the Internet but particularly at the consumer
>        edge.
> 
>     it reads like "this problem" refers to "efforts to deploy AQM" and
>     not to "increase in buffer pool size"
> 
>     ---
> 
>     (nit)
> 
>        To understand
>        queue management, it is critical to understand the difference between
>        the necessary, useful *"good" queue*, and the counterproductive *"bad" queue*.
> 
>     strictly speaking "good" and "bad" queue were not defined either
>     before or in this sentence (it is explained in section 3 or in
>     [CODEL2012])
> 
>     ---
> 
>     (nit)
> 
>        o  treat "good queue" and "bad queue" differently, that is, keep
>           delay low while permitting necessary bursts of traffic
> 
>     (1) see above
>     (2) I would say that the goal is "to keep delay low while ..." and
>     to treat good/bad queue differently is a mechanism used to achieve
>     the goal
> 
>     ---
> 
>     section 2:
> 
>     (nit)
>           In this document, *the characters ">>" preceding an indented
>     line(s)*
>        indicates a compliance requirement statement using the key words
>        listed above.  This convention aids reviewers in quickly identifying
>        or finding the explicit compliance requirements of this RFC.
> 
>     I maybe missing something, but I can't find any use of "characters
>     ">>"" anywhere except this line in any (txt,html, pdf) of the versions.
> 
>     ---
> 
>     section 3:
> 
>     generally IMHO it would be nice to introduce /interval/ and /target/
>     so it is more obvious that these are parameters (or some other kind
>     of variables). Interval is more or less obvious (although imho
>     should be more explicit). The target is first introduced at the end
>     of section 3.2 as:
> 
>        This results in a particularly simple form for the
>        setpoint: the ideal range for the permitted standing queue is between
>        5% and 10% of the TCP connection's RTT.  Thus *target* is simply 5% of
>        the interval of section 3.1
>     <https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-3.1>.
> 
>     It was not previously defined that /target/ is a parameter with a
>     meaning of /setpoint/. I don't think that this is clear for people
>     who do not know what target is in advance.
> 
>     [It seems that it was intended to introduce /interval/ and /target/
>     parameters in section 4]
> 
>     ---
> 
>        Instead of averages
>        *we recommend* tracking the minimum sojourn time, then, 
> 
>     "we recommend" or "CoDeL does"? According to the last paragraph in
>     Section 3.1, tracking minimum is a core part of CoDeL:
> 
>        These two innovations, use of sojourn time as observed values *and the local minimum as the statistic to monitor queue congestion* are key to
>        CoDel's estimator building block.
> 
>     ---
> 
>        Since the peak queue delay*is simply f r*, power is solely a function
> 
>     is it f/r ?
> 
>     ---
>     (nit)
> 
>        The only remaining building block needed for a basic AQM is a
>        'control loop' algorithm to effectively drive the queueing system
>        from any *'persistent queue above target'* state to a state where the
>        *persistent queue is below target*.
> 
>     It would be nice to explicitly state that the "persistent queue
>     above target" is indication of "bad" queue/congestion somewhere
>     above or in this line. I think that these terms are used in
>     [codel2012] but were not used before in the document. See also the
>     next item.
> 
>     ---
>     In section 3.3 there are three "states":
> 
>       *
> 
>         'persistent queue above target' state
> 
>       *
> 
>         'has persistent queue' state
> 
>       *
> 
>          dropping state
> 
>     It would be nice to a) use the same one b) explain how CoDeL gets
>     in/out of there
> 
>     ---
> 
>     (probably nit)
> 
>        When
>        *the minimum sojourn time first crosses the target* and *CoDel drops a packet*, the earliest the controller could see the effect of the drop
> 
>     from current sentence it could be inferred that the two events occur
>     at the same time which is not the case. Maybe " when persistent
>     queue above target is detected" would be better.
> 
>     ---
> 
>        variation is less than the target, and so the *initial drop spacing SHOULD be set to the estimator's interval plus
>     twice the target (i.e., initial drop spacing = 1.1 * interval)* to ensure that the
> 
>     this contradicts the pseudocode:
> 
>        dodequeue_result codel_queue_t::dodequeue(time_t now)
> 
>     	...
>     	first_above_time_ = now + interval_;
> 
>     section 4:
> 
>        To keep from
>        making drops when it would starve the output link, CoDel makes
>        another check before dropping to see if at least an MTU worth of
>        bytes remains in the buffer. If not, the packet SHOULD NOT be
>        dropped and, *currently*, CoDel exits the drop state. 
> 
>     what is the meaning of "currently" ? in the current experimental
>     version?
> 
>     ---
>     (nit)
> 
>        In practice, this value is not known or measured *(*though see
>        Section 6.2
>     <https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-6.2> for an application where interval is measured.
> 
>     missing closing bracket
> 
>     Section 5 / pseudocode: (all nits)
> 
>     The variables are introduced as first_time_above, dropping, but are
>     used as first_time_above_, dropping_ (with underscores at end). 
> 
>     ---
> 
>     *Packet** CoDelQueue::dequeue()
> 
>     a pointer to current packet was introduced as packet_t
> 
>     *--- double* now = clock();;
> 
>     time units were supposed to be of type time_t
> 
>     *--- dodequeueResult* r = dodequeue(now);
> 
>     compare with: typedef struct { ... } *dodequeue_result* below
>     (present in two routines)
> 
>     ---
> 
>     dodequeueResult r = {*NULL, queue_t::dequeue() *};
> 
>     shouldn't it be {queue_t::dequeue(), false} ?
> 
> 
>     Best Regards,
>     Polina
> 
> 
>     * note the reply by Dave Taht:
>     http://www.ietf.org/mail-archive/web/aqm/current/msg01687.html
> 
>     On 03/20/2016 04:35 PM, Wesley Eddy wrote:
>>     It looks like the WGLC feedback on the document body is
>>     incorporated, so this is good.
>>
>>     Is there a reason to stay with Informational and not Experimental
>>     like we've done with PIE an d FQ-CoDel?
>>
>>     Also, idnits has some problems with the references that should be
>>     fixed (e.g. "NATAL2010" is probably supposed to be "NETAL2010"):
>>     https://tools.ietf.org/idnits?url=https://tools.ietf.org/id/draft-ietf-aqm-codel-03.txt
> 
> 
>     _______________________________________________
>     aqm mailing list
>     aqm@ietf.org <mailto:aqm@ietf.org>
>     https://www.ietf.org/mailman/listinfo/aqm
> 
> 
> 
> 
> -- 
> Andrew McGregor | SRE | andrewmcgr@google.com
> <mailto:andrewmcgr@google.com> | +61 4 1071 2221
> 
> 
> _______________________________________________
> aqm mailing list
> aqm@ietf.org
> https://www.ietf.org/mailman/listinfo/aqm
> 

-- 
Prof. Dr. habil. Michael Menth
University of Tuebingen
Faculty of Science
Department of Computer Science
Chair of Communication Networks
Sand 13, 72076 Tuebingen, Germany
phone: (+49)-7071/29-70505
fax: (+49)-7071/29-5220
mailto:menth@uni-tuebingen.de
http://kn.inf.uni-tuebingen.de


From nobody Thu Jul  7 06:54:37 2016
Return-Path: <chromatix99@gmail.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 4B42D12D761 for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 06:54:37 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -1.45
X-Spam-Level: 
X-Spam-Status: No, score=-1.45 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, FSL_HELO_HOME=1, RCVD_IN_DNSWL_LOW=-0.7, SPF_PASS=-0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
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 va_mEQjPJXuQ for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 06:54:35 -0700 (PDT)
Received: from mail-lf0-x243.google.com (mail-lf0-x243.google.com [IPv6:2a00:1450:4010:c07::243]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 4A61812D09F for <aqm@ietf.org>; Thu,  7 Jul 2016 06:54:35 -0700 (PDT)
Received: by mail-lf0-x243.google.com with SMTP id l188so1523925lfe.0 for <aqm@ietf.org>; Thu, 07 Jul 2016 06:54:35 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113;  h=mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=xjAkjsQpm5hitYsNjlnlwAGg8I1I//U5/WNbOQAH7C0=; b=yaSEK0hE1qMjY0FOOXD4tJ7Y2CtbN51Ke78fEDrYNaxe3+p0baqP37CQo257KmMdi7 8MbE70dG9cf2jViZUpMprTueOXldPP0ZEv0PcVndnApvjbK0iIU/xaXWOGFDda+DTBik pxatf8sUBvBv65WdPmQB09H/OcjMXcdw3mG8Uw2E0giRfVuVnWnxH8MThELwMqIEOBuI ehI9G8rjxc9Lo8QG5dGvNBuNY9+lU79Bkoz0Gy4Xkv0yY/Tz3X8bQ/ZryAKmz+5SvUJ4 QwnOPVnwLYHg+8/vOIgeUxUo439jmwbx6dhjc+yt+gS4f4k/yPEKEWF+yzMzpgQeujgJ h15g==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=xjAkjsQpm5hitYsNjlnlwAGg8I1I//U5/WNbOQAH7C0=; b=ZhqYfCYhmi6Huvji81THe9iL8GWR4hjcxfoGQFVOwmpMGMR3ATeeZstJAUMz9eCCLr 4MH57zNcyVXK7+BtM5iBpzaR5QcP3xuWoCKDeF4ifECWwKds1yAPgSeE1eOwpPdqzMb/ ocBGgyRY+AJcOVvLRlqpxci3+zNPt0xH/Sz/R8TyUmOi+MqVMwg8QqpCGy2xtlVw436w /OYAjEv2uacBxBEnJh4j8ivBFnPD40yPjXc1LJhFR8MokJKzXkioxGpRmbFIrSnywkIn dFUzE1U/t4kHSb1oQUSyv80OKjvVnT8CdlhpllWEe7Ncn+pW+hecqYUO6KOkrnbDbduS Eq+Q==
X-Gm-Message-State: ALyK8tILZWrdorB3RbwS+dpKXuWyHvSYFd2RsSkXZF9O0TmX9Z65HAga8yUzfELzwuQ8rQ==
X-Received: by 10.46.1.27 with SMTP id 27mr102603ljb.26.1467899673410; Thu, 07 Jul 2016 06:54:33 -0700 (PDT)
Received: from bass.home.chromatix.fi (37-33-96-207.bb.dnainternet.fi. [37.33.96.207]) by smtp.gmail.com with ESMTPSA id g16sm994742lji.44.2016.07.07.06.54.31 (version=TLS1 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Thu, 07 Jul 2016 06:54:32 -0700 (PDT)
Content-Type: text/plain; charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
From: Jonathan Morton <chromatix99@gmail.com>
In-Reply-To: <577E4C53.9030909@uni-tuebingen.de>
Date: Thu, 7 Jul 2016 16:54:28 +0300
Content-Transfer-Encoding: quoted-printable
Message-Id: <88A51A3A-64EE-4A6C-99A5-7EA8DC33CC8B@gmail.com>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu> <577E4C53.9030909@uni-tuebingen.de>
To: Michael Menth <menth@uni-tuebingen.de>
X-Mailer: Apple Mail (2.3124)
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/6JH6qNdkydfWKHfdedX6zw3i2_E>
Cc: "Goltsman, Polina" <polina.goltsman@student.kit.edu>, Andrew Mcgregor <andrewmcgr@google.com>, Fabian Schwarzkopf <f.schwarzkopf@yahoo.de>, Wesley Eddy <wes@mti-systems.com>, Dave Taht <dave.taht@gmail.com>, Jana Iyengar <jri@google.com>, "Bless, Roland \(TM\)" <roland.bless@kit.edu>, "aqm@ietf.org" <aqm@ietf.org>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 13:54:37 -0000

> On 7 Jul, 2016, at 15:34, Michael Menth <menth@uni-tuebingen.de> =
wrote:
>=20
>> Based on our evaluations, with pure CoDel (without FQ-CoDel),
>> "reentering" is actually a common case. I think Dave and Toke should
>> have more experimental results to answer this question. (I included =
Dave
>> in CC)
>=20
> We also studied this issue about one or two years ago. The algorithm
> controlling the reentering of the dropping mode may have a significant
> impact on control variables, drop rates, drop pattern depending on the
> traffic. It can also impact utilization. Most interesting is probably
> Fig. 6 on page 7 in
> https://atlas.informatik.uni-tuebingen.de/~menth/papers/Menth16e.pdf

That=E2=80=99s a very interesting paper for Codel enthusiasts.  IMHO, =
Codel-ACT performs much better on the important metrics than plain =
Codel, especially when faced with multiplexed flows.

I had also noticed the potential for Codel=E2=80=99s count to increase =
unboundedly while working on Cake; in fact when implemented as a 16-bit =
counter, it would wrap around on certain workloads, causing apparently =
chaotic behaviour.  I therefore modified the variant of Codel that Cake =
used to *halve* the count on re-entry, a much more aggressive reduction =
than Codel-ACT uses, as well as using a saturating increment to prevent =
wraparound.  However, I kept the basic behaviour of waiting a full =
interval before re-entry.

More recently, I have changed Cake over to use COBALT, which is a =
combination of Codel and BLUE; Codel is intended to handle responsive =
flows which behave in a TCP-like manner, and BLUE to take over if the =
flow proves to be unresponsive (by running into the queue length limit).

COBALT uses a re-implemented version of Codel with a much simplified =
main loop.  It also has a novel re-entry behaviour: the count is reduced =
while the queue length is below target by running the algorithm in =
reverse, without marking or dropping.  This means that very short =
non-dropping periods result in a slight reduction in count, and longer =
periods result in a larger reduction; the reduction also occurs faster =
when count is high.

With your paper in mind, I may also change it to use the shorter =
re-entry delay of Codel-ACT, as you explain very clearly how it may be =
beneficial, and this shows in the graphs as a lower peak delay and a =
reduction in loss synchronisation between flows.  I also think it will =
allow Codel to adapt more cleanly to RTTs significantly shorter than it =
is configured for - this is important where the defaults are set for =
internet-scale traffic, and people unthinkingly rely on those defaults =
in a LAN environment.

Here is the relevant part of COBALT=E2=80=99s Codel implementation, =
showing the re-entry count reduction strategy:

if(next_due && vars->dropping) {
	/* Use ECN mark if possible, otherwise drop */
	drop =3D !(vars->ecn_marked =3D INET_ECN_set_ce(skb));

	vars->count++;
	if(!vars->count)
		vars->count--;
	cobalt_invsqrt(vars);
	vars->drop_next =3D cobalt_control_law(vars->drop_next, =
p->interval, vars->rec_inv_sqrt);
} else {
	while(next_due) {
		vars->count--;
		cobalt_invsqrt(vars);
		vars->drop_next =3D cobalt_control_law(vars->drop_next, =
p->interval, vars->rec_inv_sqrt);
		schedule =3D now - vars->drop_next;
		next_due =3D vars->count && schedule >=3D 0;
	}
}

The full code can be found here:
	https://github.com/dtaht/sch_cake/blob/master/cobalt.c

 - Jonathan Morton


From nobody Thu Jul  7 12:04:18 2016
Return-Path: <dave@taht.net>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id DC34F12B059 for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:04:16 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.327
X-Spam-Level: 
X-Spam-Status: No, score=-3.327 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RP_MATCHES_RCVD=-1.426, SPF_PASS=-0.001] autolearn=ham autolearn_force=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 LhiV2kIjJdQn for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:04:13 -0700 (PDT)
Received: from mail.taht.net (mail.taht.net [176.58.107.8]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 85D0712B019 for <aqm@ietf.org>; Thu,  7 Jul 2016 12:04:13 -0700 (PDT)
Received: from dair-2430.local (cpc5-hawk13-2-0-cust174.aztw.cable.virginm.net [82.45.7.175]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by mail.taht.net (Postfix) with ESMTPSA id 4B62221316; Thu,  7 Jul 2016 19:04:10 +0000 (UTC)
To: aqm@ietf.org, "codel@lists.bufferbloat.net" <codel@lists.bufferbloat.net>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu> <577E4C53.9030909@uni-tuebingen.de>
From: =?UTF-8?Q?Dave_T=c3=a4ht?= <dave@taht.net>
Message-ID: <cf19993f-3c00-f171-f59a-f76692d0da14@taht.net>
Date: Thu, 7 Jul 2016 20:11:55 +0100
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.9; rv:45.0) Gecko/20100101 Thunderbird/45.1.1
MIME-Version: 1.0
In-Reply-To: <577E4C53.9030909@uni-tuebingen.de>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 8bit
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/cxjo3rNxqb6y0VVoM_qrBQUbKCk>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 19:04:17 -0000

On 7/7/16 1:34 PM, Michael Menth wrote:
> Dear Polina, dear all,
> 
> Am 07.07.2016 um 13:49 schrieb Goltsman, Polina:
>> Dear Andrew, Dear Wesley,
>>
>>
>> First of all, sorry for the long waiting time. I still haven't processed
>> "nits" but I have 2 major issues.
>>
>>
>> *Issue 1: *This issue is still related to the reentering dropping state.
>>
>>
>> As of the latest I-D, the "reentering dropping state"  appears in form
>> of one sentence in the text and a small code block in the pseudo-code.
>>
>>
>> The text is on page 13:
>>
>>> Additional logic prevents re-
>>> entering the dropping state too soon after exiting it and resumes the
>>> dropping state at a recent control level, if one exists.
>>
>> The comment to the pseudo-code on page 20:
>>
>>
>>            // If min went above target close to when it last went
>>            // below, assume that the drop rate that controlled the
>>            // queue on the last cycle is a good starting point to
>>            // control it now. ('drop_next' will be at most 'interval'
>>            // later than the time of the last drop so 'now - drop_next'
>>            // is a good approximation of the time from the last drop
>>            // until now.) *Implementations vary slightly here; this is // the Linux version, which
>> is more widely deployed and // tested.
>>
>> *
>>
>>
>> Or summarized: this issue is not addressed in text, which instead just
>> mentions that the issue exists, there are several versions of this logic
>> and here is one of them.
>>
>>
>> As it reads now, this "reentering" should be a tiny insignificant part
>> of the algorithm that (1) doesn't happen in "most" of the cases, and (2)
>> when it does happen, it will only "slightly" affect "performance" (with
>> proper definitions of "most", "slightly", and "performance").
>>
>>
>> Is it really so? 
>>
>>
>> Based on our evaluations, with pure CoDel (without FQ-CoDel),
>> "reentering" is actually a common case. I think Dave and Toke should
>> have more experimental results to answer this question. 


It is actually a common case. The initial run-up of the count variable
dominates in the early stages of a many-TCP onslaught, then later on it
falls into and out of the drop state portion in the latter part of the
code, being discussed here.

Flent has suport for sampling the delay, drops, and queue backlogs, and
we have data for 1-1000 flows, <1 to 100s ms of rtts through bottleneck
links ranging from 1mbit through a gbit.

We have not been sampling the value of the count variable in those
tests. Most of our testing has been limited to persistent loads of 20
minutes or less, and has tended to focus on bidirectional and web
traffic, using fq_codel derivatives, rather than pounding X number of
full rate flows through the system.

>>(I included Dave in CC)

I am on the aqm, bloat, and codel lists, but am traveling through the EU
(Well, england just now), through mid-august, so replies may be
infrequent and/or incomplete. I will be at ietf.

> 
> We also studied this issue about one or two years ago. The algorithm
> controlling the reentering of the dropping mode may have a significant
> impact on control variables, drop rates, drop pattern depending on the
> traffic. It can also impact utilization. Most interesting is probably
> Fig. 6 on page 7 in
> https://atlas.informatik.uni-tuebingen.de/~menth/papers/Menth16e.pdf

A couple quick, scattered notes:

re:


I have never been happy with the resume/decay portion of the algorithm,
and thank you for exploring some of the alternatives left behind on the
simulation bench.

I'd have liked the linux version of this bit plugged into the experiment.

Floating point is unavailable in the linux kernel, and while we could do
a fixed point fractional count, getting up to the number of colliding
flows needed to start needing that in fq_codel seemed not needed.

Adding the -ACT ideas to fq_codel would be an interesting test.
I think, from reading this paper, that we could easily recreate the
conditions of this experiment in flent, but I would always welcome
source code so as to be able to change variables like the kernel tcp
version from kernel 2.6.21 to 4.8, factor in BQL, look at RTTs between
10 and 100, bandwidths > 10mbit, and other underlying issues like that.



In the 64 flows case at 10mbit, with the sustained low delay of
codel-act, and pie, I am curious how often tcp flows are resetting or
having RTOs? I do not think that a fixed delay is desirable. "You aim
for a target", not that you need to hit it. More important is high
utilization, and trying to keep the connections afloat, with as minimum
latency as possible.

I would have liked to seen a staggered start attempted for this many
flows, reverse traffic, etc.

> Kind regards,
> 
> Michael
> 
> 
> 
>>
>>
>> If it is so, I still think that the "common case" should be at the very
>> least explained in text too.
>>
>>
>> *Issue 2:* The "initial drop spacing" is inconsistent
>>
>>
>> Regarding the "initial drop spacing" there are following lines in the
>> document.
>>
>>
>> between pages 11 and 12:
>>
>>> and so the initial drop spacing 
>>> SHOULD be set to the estimator's *interval plus twice the target*
>>> (i.e., initial drop spacing = 1.1 * interval)
>>
>> page 12:
>>
>>> the initial next
>>> drop spacing is intended to be long enough to give the endpoints time
>>> to react to the single drop so SHOULD be set to a value *of 1.1 times >the interval*
>>
>> page 16:
>>
>>
>>> and the initial drop
>>> spacing is also set to *interval*.
>>
>> and it is also set to*one interval* in pseudo-code, which I believe is
>> this line:
>>
>>
>>> drop_next_ = control_law(now, count_);
>>
>>
>> Note: I think there was an email on CoDel mailing list
>> (https://lists.bufferbloat.net/pipermail/codel/) about this issue.
>>
>>
>>
>> ------------------------------------------------------------------------
>> *From:* Andrew Mcgregor <andrewmcgr@google.com>
>> *Sent:* Friday, June 3, 2016 3:50 AM
>> *To:* Goltsman, Polina
>> *Cc:* Wesley Eddy; aqm@ietf.org; Jana Iyengar
>> *Subject:* Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
>>  
>> Thank you Polina.
>>
>> We have incorporated most of these (sometimes with slightly different
>> solutions than you suggested), and a new draft will be out very
>> shortly.  I felt that by far the most important was the different way
>> the Linux implementation addressed re-entry to the dropping state, and
>> the pseudocode now does exactly that.
>>
>> Andrew
>>
>> On 22 March 2016 at 08:10, Polina Goltsman
>> <polina.goltsman@student.kit.edu
>> <mailto:polina.goltsman@student.kit.edu>> wrote:
>>
>>     Dear Wesley, Dear All,
>>
>>     First of all our feedback regarding different "re-entering dropping
>>     state" in the document and in the Linux implementation
>>     (http://www.ietf.org/mail-archive/web/aqm/current/msg01686.html) was
>>     not addressed.
>>
>>     As FQ-CoDel  relies on CoDel, this issue is also (partly) relevant
>>     for the FQ-CoDel document. In the introduction FQ-CoDel references
>>     ns-3 and Linux implementations where the first one uses the
>>     re-entering logic from the CoDel document while the second from
>>     CoDel Linux implementation. The algorithm that has seen widespread
>>     testing according to Section 7 is (I suppose) the Linux version. Is
>>     this situation acceptable for an algorithm specification?
>>
>>     /[since this comment was supposed to be sent before the end of 2015,
>>     feel free to (silently) ignore it]/
>>     Second, unlike Rasool Al-Saadi (see
>>     http://www.ietf.org/mail-archive/web/aqm/current/msg01693.html) I do
>>     not like the document. Although I agree that the pseudocode is
>>     sufficient to create a working implementation, however, in my
>>     opinion, the rest of the document makes implementing CoDeL more
>>     confusing (at least without reading [CODEL2012] first). Is it normal
>>     for a RFC, which, as I assume, should primarily contain an algorithm
>>     specification to contain the algorithm specification ONLY in form of
>>     pseudocode?  
>>     These are two items that I found the most confusing:
>>     (1) section 3 introduces estimator, setpoint, and control loop which
>>     are not clearly distinguishable in pseudocode. It would be nice if
>>     section 4 explained how the three entities transition into the
>>     routines in pseudocode.
>>     (2) IMHO, there is some missing explanations. For example, the
>>     document never says how exactly "bad"/persistent queue is
>>     determined. The document says in Section 3.1 that the minimum is
>>     tracked, but it never says that persistent queue is when the minimum
>>     is above setpoint/target for at least an interval. As another
>>     example, Section 3.3 could say how the controller looks like exactly.
>>
>>     I also have some small comments and nits regarding the latest
>>     version (-03):
>>
>>     ---
>>
>>     in the whole document there are several places of using "we" to
>>     (supposedly) refer to the authors. According to Bob Briscoe's
>>     message
>>     (http://www.ietf.org/mail-archive/web/aqm/current/msg01805.html),
>>     "we" do not necessary refer to the authors in a rfc. Are these "we"s ok?
>>
>>     ---
>>
>>     abstract:
>>
>>     This document describes a general *framework* called CoDel 
>>
>>     IMHO the word "framework" is too overloaded and too broad. It might
>>     be better to call CoDel an AQM or something that has "queue" in it.
>>     Other ways (e.g., http://queue.acm.org/detail.cfm?id=2839461)
>>     probably also use it to control a queue.
>>
>>     ---
>>
>>     section 1:
>>
>>     (nit)
>>        Despite this
>>        awareness, the problem has only gotten worse as Moore's Law growth in
>>        memory density fueled an exponential increase in buffer pool size.
>>        Efforts to deploy AQM have been frustrated by difficult configuration
>>        and negative impact on network utilization.  *This problem,* recently
>>        christened "bufferbloat", [TSV2011] [BB2011] has become increasingly
>>        important throughout the Internet but particularly at the consumer
>>        edge.
>>
>>     it reads like "this problem" refers to "efforts to deploy AQM" and
>>     not to "increase in buffer pool size"
>>
>>     ---
>>
>>     (nit)
>>
>>        To understand
>>        queue management, it is critical to understand the difference between
>>        the necessary, useful *"good" queue*, and the counterproductive *"bad" queue*.
>>
>>     strictly speaking "good" and "bad" queue were not defined either
>>     before or in this sentence (it is explained in section 3 or in
>>     [CODEL2012])
>>
>>     ---
>>
>>     (nit)
>>
>>        o  treat "good queue" and "bad queue" differently, that is, keep
>>           delay low while permitting necessary bursts of traffic
>>
>>     (1) see above
>>     (2) I would say that the goal is "to keep delay low while ..." and
>>     to treat good/bad queue differently is a mechanism used to achieve
>>     the goal
>>
>>     ---
>>
>>     section 2:
>>
>>     (nit)
>>           In this document, *the characters ">>" preceding an indented
>>     line(s)*
>>        indicates a compliance requirement statement using the key words
>>        listed above.  This convention aids reviewers in quickly identifying
>>        or finding the explicit compliance requirements of this RFC.
>>
>>     I maybe missing something, but I can't find any use of "characters
>>     ">>"" anywhere except this line in any (txt,html, pdf) of the versions.
>>
>>     ---
>>
>>     section 3:
>>
>>     generally IMHO it would be nice to introduce /interval/ and /target/
>>     so it is more obvious that these are parameters (or some other kind
>>     of variables). Interval is more or less obvious (although imho
>>     should be more explicit). The target is first introduced at the end
>>     of section 3.2 as:
>>
>>        This results in a particularly simple form for the
>>        setpoint: the ideal range for the permitted standing queue is between
>>        5% and 10% of the TCP connection's RTT.  Thus *target* is simply 5% of
>>        the interval of section 3.1
>>     <https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-3.1>.
>>
>>     It was not previously defined that /target/ is a parameter with a
>>     meaning of /setpoint/. I don't think that this is clear for people
>>     who do not know what target is in advance.
>>
>>     [It seems that it was intended to introduce /interval/ and /target/
>>     parameters in section 4]
>>
>>     ---
>>
>>        Instead of averages
>>        *we recommend* tracking the minimum sojourn time, then, 
>>
>>     "we recommend" or "CoDeL does"? According to the last paragraph in
>>     Section 3.1, tracking minimum is a core part of CoDeL:
>>
>>        These two innovations, use of sojourn time as observed values *and the local minimum as the statistic to monitor queue congestion* are key to
>>        CoDel's estimator building block.
>>
>>     ---
>>
>>        Since the peak queue delay*is simply f r*, power is solely a function
>>
>>     is it f/r ?
>>
>>     ---
>>     (nit)
>>
>>        The only remaining building block needed for a basic AQM is a
>>        'control loop' algorithm to effectively drive the queueing system
>>        from any *'persistent queue above target'* state to a state where the
>>        *persistent queue is below target*.
>>
>>     It would be nice to explicitly state that the "persistent queue
>>     above target" is indication of "bad" queue/congestion somewhere
>>     above or in this line. I think that these terms are used in
>>     [codel2012] but were not used before in the document. See also the
>>     next item.
>>
>>     ---
>>     In section 3.3 there are three "states":
>>
>>       *
>>
>>         'persistent queue above target' state
>>
>>       *
>>
>>         'has persistent queue' state
>>
>>       *
>>
>>          dropping state
>>
>>     It would be nice to a) use the same one b) explain how CoDeL gets
>>     in/out of there
>>
>>     ---
>>
>>     (probably nit)
>>
>>        When
>>        *the minimum sojourn time first crosses the target* and *CoDel drops a packet*, the earliest the controller could see the effect of the drop
>>
>>     from current sentence it could be inferred that the two events occur
>>     at the same time which is not the case. Maybe " when persistent
>>     queue above target is detected" would be better.
>>
>>     ---
>>
>>        variation is less than the target, and so the *initial drop spacing SHOULD be set to the estimator's interval plus
>>     twice the target (i.e., initial drop spacing = 1.1 * interval)* to ensure that the
>>
>>     this contradicts the pseudocode:
>>
>>        dodequeue_result codel_queue_t::dodequeue(time_t now)
>>
>>     	...
>>     	first_above_time_ = now + interval_;
>>
>>     section 4:
>>
>>        To keep from
>>        making drops when it would starve the output link, CoDel makes
>>        another check before dropping to see if at least an MTU worth of
>>        bytes remains in the buffer. If not, the packet SHOULD NOT be
>>        dropped and, *currently*, CoDel exits the drop state. 
>>
>>     what is the meaning of "currently" ? in the current experimental
>>     version?
>>
>>     ---
>>     (nit)
>>
>>        In practice, this value is not known or measured *(*though see
>>        Section 6.2
>>     <https://tools.ietf.org/html/draft-ietf-aqm-codel-03#section-6.2> for an application where interval is measured.
>>
>>     missing closing bracket
>>
>>     Section 5 / pseudocode: (all nits)
>>
>>     The variables are introduced as first_time_above, dropping, but are
>>     used as first_time_above_, dropping_ (with underscores at end). 
>>
>>     ---
>>
>>     *Packet** CoDelQueue::dequeue()
>>
>>     a pointer to current packet was introduced as packet_t
>>
>>     *--- double* now = clock();;
>>
>>     time units were supposed to be of type time_t
>>
>>     *--- dodequeueResult* r = dodequeue(now);
>>
>>     compare with: typedef struct { ... } *dodequeue_result* below
>>     (present in two routines)
>>
>>     ---
>>
>>     dodequeueResult r = {*NULL, queue_t::dequeue() *};
>>
>>     shouldn't it be {queue_t::dequeue(), false} ?
>>
>>
>>     Best Regards,
>>     Polina
>>
>>
>>     * note the reply by Dave Taht:
>>     http://www.ietf.org/mail-archive/web/aqm/current/msg01687.html
>>
>>     On 03/20/2016 04:35 PM, Wesley Eddy wrote:
>>>     It looks like the WGLC feedback on the document body is
>>>     incorporated, so this is good.
>>>
>>>     Is there a reason to stay with Informational and not Experimental
>>>     like we've done with PIE an d FQ-CoDel?
>>>
>>>     Also, idnits has some problems with the references that should be
>>>     fixed (e.g. "NATAL2010" is probably supposed to be "NETAL2010"):
>>>     https://tools.ietf.org/idnits?url=https://tools.ietf.org/id/draft-ietf-aqm-codel-03.txt
>>
>>
>>     _______________________________________________
>>     aqm mailing list
>>     aqm@ietf.org <mailto:aqm@ietf.org>
>>     https://www.ietf.org/mailman/listinfo/aqm
>>
>>
>>
>>
>> -- 
>> Andrew McGregor | SRE | andrewmcgr@google.com
>> <mailto:andrewmcgr@google.com> | +61 4 1071 2221
>>
>>
>> _______________________________________________
>> aqm mailing list
>> aqm@ietf.org
>> https://www.ietf.org/mailman/listinfo/aqm
>>
> 


From Carl_Klatsky@comcast.com  Thu Jul  7 12:12:51 2016
Return-Path: <Carl_Klatsky@comcast.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id 9B1B912D0AC for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:12:51 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -3.327
X-Spam-Level: 
X-Spam-Status: No, score=-3.327 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RP_MATCHES_RCVD=-1.426, SPF_PASS=-0.001] autolearn=ham autolearn_force=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 P0L0EENfe9gw for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:12:49 -0700 (PDT)
Received: from pacdcmhout01.cable.comcast.com (PACDCMHOUT01.cable.comcast.com [68.87.31.167]) (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 78A6112B068 for <aqm@ietf.org>; Thu,  7 Jul 2016 12:12:49 -0700 (PDT)
X-AuditID: 44571fa7-fa76f98000001305-48-577ea9b064d5
Received: from PACDCEX48.cable.comcast.com (cas-umc02.ndceast.pa.bo.comcast.net [68.87.34.28]) (using TLS with cipher AES256-SHA256 (256/256 bits)) (Client did not present a certificate) by  (SMTP Gateway) with SMTP id 40.0F.04869.0B9AE775; Thu,  7 Jul 2016 15:12:48 -0400 (EDT)
Received: from PACDCEX43.cable.comcast.com (24.40.2.142) by PACDCEX48.cable.comcast.com (24.40.2.147) with Microsoft SMTP Server (TLS) id 15.0.1130.7; Thu, 7 Jul 2016 15:12:46 -0400
Received: from PACDCEX43.cable.comcast.com ([fe80::3aea:a7ff:fe36:87a8]) by PACDCEX43.cable.comcast.com ([fe80::3aea:a7ff:fe36:87a8%19]) with mapi id 15.00.1130.005; Thu, 7 Jul 2016 15:12:46 -0400
From: "Klatsky, Carl" <Carl_Klatsky@comcast.com>
To: Jonathan Morton <chromatix99@gmail.com>, Michael Menth <menth@uni-tuebingen.de>
Thread-Topic: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
Thread-Index: AQHRfxb87A5LHxhy8060NvHhQsbonJ9iwUKAgAHv6ACAcwh1AIA2FogAgAAMhYCAABZcAIAAArgw
Date: Thu, 7 Jul 2016 19:12:45 +0000
Message-ID: <7ad383c30696497dae5dfd89ded5274f@PACDCEX43.cable.comcast.com>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu> <577E4C53.9030909@uni-tuebingen.de> <88A51A3A-64EE-4A6C-99A5-7EA8DC33CC8B@gmail.com>
In-Reply-To: <88A51A3A-64EE-4A6C-99A5-7EA8DC33CC8B@gmail.com>
Accept-Language: en-US
Content-Language: en-US
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
x-ms-exchange-transport-fromentityheader: Hosted
x-originating-ip: [68.87.29.10]
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: base64
MIME-Version: 1.0
X-CFilter-Loop: Forward
X-Brightmail-Tracker: H4sIAAAAAAAAA+NgFnrBIsWRmVeSWpSXmKPExsXiEq4ko7thZV24QfseFYveY/PYLNbsk7RY uO0Ho8Wv70vYLCYdXchi8ebIWhaLxRM62Cw+TZjOanH9O7sDp8fOWXfZPRZsKvVYsuQnk8eC hTuYPU6e6mXz6Nrbwu7Rv/8hu8eEG48ZAziiuGxSUnMyy1KL9O0SuDIu/D/HVPBCrWLngXls DYwvVLsYOTkkBEwkDm9YwtbFyMUhJDCHSaLz53soZyejxL09N6GcE4wSn891MIG0sAkYSPy6 +5UFxBYRCJX4O+UAO0gRs8BxJonLE1+xgySEBawk1n/awwpRZC3Rt3oLG4QdJfGiYTZYDYuA isT3ufcZQWxeAS+Jhp2bmSG2nWOSWLh6BlARBwengK3EvEt2IDWMAmIS30+tATuCWUBc4taT +UwQPwhILNlznhnCFpV4+fgfK4RtILF16T4WCFte4siEfywgI5kFNCXW79KHGKMoMaX7ITvE CYISJ2c+gSoXlzh8ZAfrBEaJWUi2zULonoWkexaS7gWMLKsY5QoSk1OSczPyS0sMDPWSE5Ny UvWS83OTE4tLQPQmRnD8yy/fwXjvhdMhRgEORiUeXovWunAh1sSy4srcQ4wSHMxKIrycy4FC vCmJlVWpRfnxRaU5qcWHGKU5WJTEeV9NigkXEkhPLEnNTk0tSC2CyTJxcEo1MAb9ZHz895wP T9e8JjOmwMclR516vsyznLku0mkKv5z7ty0Prkq2/H2+VrqAJc5bY8bivIuBMYd5rdbr6Qpf Ep00w3D6+blZSUdibE54TJ2+RmzX0/3FLdnTIuU+fJDQcD2cJin75Wg9S7fzHIatHPxH3yca n1QUS/4702OKaP9X8cnyxYHH+pRYijMSDbWYi4oTAXquyHf7AgAA
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/Io2hnrk5S89YLQEHBbpajMV3lN8>
X-Mailman-Approved-At: Thu, 07 Jul 2016 12:23:28 -0700
Cc: "Goltsman, Polina" <polina.goltsman@student.kit.edu>, Andrew Mcgregor <andrewmcgr@google.com>, Fabian Schwarzkopf <f.schwarzkopf@yahoo.de>, Wesley Eddy <wes@mti-systems.com>, Dave Taht <dave.taht@gmail.com>, Jana Iyengar <jri@google.com>, "Bless, Roland \(TM\)" <roland.bless@kit.edu>, "aqm@ietf.org" <aqm@ietf.org>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 19:13:58 -0000
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From nobody Thu Jul  7 12:23:33 2016
Return-Path: <chromatix99@gmail.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id A6A9812D1CB for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:15:15 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -0.75
X-Spam-Level: 
X-Spam-Status: No, score=-0.75 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FREEMAIL_ENVFROM_END_DIGIT=0.25, FREEMAIL_FROM=0.001, FSL_HELO_HOME=1, SPF_PASS=-0.001] autolearn=no autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
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 ZFyOq_0zVwc1 for <aqm@ietfa.amsl.com>; Thu,  7 Jul 2016 12:15:14 -0700 (PDT)
Received: from mail-lf0-x229.google.com (mail-lf0-x229.google.com [IPv6:2a00:1450:4010:c07::229]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id DF46112D0C2 for <aqm@ietf.org>; Thu,  7 Jul 2016 12:15:13 -0700 (PDT)
Received: by mail-lf0-x229.google.com with SMTP id h129so17687174lfh.1 for <aqm@ietf.org>; Thu, 07 Jul 2016 12:15:13 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113;  h=mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=vBEtZOLbt3OZhadxUFN9Bn5E48QVLSv6Tu0rACQc+qE=; b=bJgb40utCaaVjThFnvwY9/NXZ7Py4tOxnfLHudQWdzfDPnBpDzZ5T4IOLxDSJm9erw odOopB2weFFRzuuYlRu7MFB6TRISjAP6B2wPfn/5PkJuceCnuMgLifsqCgNq7T1mbeQA 0IBLeXpjCdy/vxOGq8Y3xKPlUEegoYmUKHqqAJ/e65UXHWfKL+BPn9yYP7HLcqSJJmb5 9FMODhjIuf5GzkgHjVKt3fCbYzMvsBNrSjoEcRgkIHzYHMHg2JxJzp2y57dmRhdzZJiV e4dzYts8HvSMpD+YTpbmubrRkqE9dY9rzG3LOIARY3LAoUEaCuFOc5kyGrNlpphoS9Rm gKRg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:subject:from:in-reply-to:date:cc :content-transfer-encoding:message-id:references:to; bh=vBEtZOLbt3OZhadxUFN9Bn5E48QVLSv6Tu0rACQc+qE=; b=lNzjXHaK6lnpBGpHp9aXGMyetKjDhClQ/fIOKwVbz6F0LNyzxu8CVU4eqLXRZykI22 OUYb80GWXM20oE1EKlQQpSfh5ppGv2hTIUyn+WTjDUZ84uWcZo0FM753Xrh+aZKvMYTj +vbIRWoixz0m8Qe+mpVMR+rSlGkiW1OJtUStwCYLhuNLPtxylf31a8p8ndLVmv30ejPF Y37U5fZ7yVFA0niQPwmW15kFq39dVSx0S1jCwju3QlRdwDNIqu83RSW1PGv00f6yLSi2 P+7ULAr2Ke90D7Tvew9jwXgp34STnE/w6qEg30PWPouukSoQoKYwOr+4EtUGTkmLjS17 En1w==
X-Gm-Message-State: ALyK8tLg4dOXhGYNb4A6UgaQY9iWIrnTSDZQI11ERj1JUHoLrl1B1A1P4uGdUzljnaas5A==
X-Received: by 10.46.5.208 with SMTP id 199mr507389ljf.12.1467918912010; Thu, 07 Jul 2016 12:15:12 -0700 (PDT)
Received: from bass.home.chromatix.fi (37-33-96-207.bb.dnainternet.fi. [37.33.96.207]) by smtp.gmail.com with ESMTPSA id m82sm372107lfm.2.2016.07.07.12.15.10 (version=TLS1 cipher=ECDHE-RSA-AES128-SHA bits=128/128); Thu, 07 Jul 2016 12:15:11 -0700 (PDT)
Content-Type: text/plain; charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 9.3 \(3124\))
From: Jonathan Morton <chromatix99@gmail.com>
In-Reply-To: <7ad383c30696497dae5dfd89ded5274f@PACDCEX43.cable.comcast.com>
Date: Thu, 7 Jul 2016 22:15:07 +0300
Content-Transfer-Encoding: 7bit
Message-Id: <244EAEB7-8341-4402-8B9E-E888B59EEF36@gmail.com>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu> <577E4C53.9030909@uni-tuebingen.de> <88A51A3A-64EE-4A6C-99A5-7EA8DC33CC8B@gmail.com> <7ad383c30696497dae5dfd89ded5274f@PACDCEX43.cable.comcast.com>
To: "Klatsky, Carl" <Carl_Klatsky@comcast.com>
X-Mailer: Apple Mail (2.3124)
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/DRtrUju-q6DSt5sskhPEyh2R5hY>
X-Mailman-Approved-At: Thu, 07 Jul 2016 12:23:28 -0700
Cc: "Goltsman, Polina" <polina.goltsman@student.kit.edu>, Andrew Mcgregor <andrewmcgr@google.com>, Fabian Schwarzkopf <f.schwarzkopf@yahoo.de>, Wesley Eddy <wes@mti-systems.com>, Dave Taht <dave.taht@gmail.com>, Michael Menth <menth@uni-tuebingen.de>, Jana Iyengar <jri@google.com>, "Bless, Roland \(TM\)" <roland.bless@kit.edu>, "aqm@ietf.org" <aqm@ietf.org>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 07 Jul 2016 19:15:16 -0000

> On 7 Jul, 2016, at 22:12, Klatsky, Carl <Carl_Klatsky@comcast.com> wrote:
> 
>> https://atlas.informatik.uni-tuebingen.de/~menth/papers/Menth16e.pdf

> [CK] I am familiar with CoDel and FQ-CoDel, but what is CoDel-ACT?

It is described in the above-linked paper.

 - Jonathan Morton


From nobody Sun Jul 10 14:22:46 2016
Return-Path: <roland.bless@kit.edu>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id C17AD12D109 for <aqm@ietfa.amsl.com>; Sun, 10 Jul 2016 14:22:44 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.301
X-Spam-Level: 
X-Spam-Status: No, score=-2.301 tagged_above=-999 required=5 tests=[BAYES_40=-0.001, RCVD_IN_DNSWL_MED=-2.3] autolearn=ham autolearn_force=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 3FtIIl8vvtqL for <aqm@ietfa.amsl.com>; Sun, 10 Jul 2016 14:22:43 -0700 (PDT)
Received: from iramx2.ira.uni-karlsruhe.de (iramx2.ira.uni-karlsruhe.de [141.3.10.81]) (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 DF21B12D106 for <aqm@ietf.org>; Sun, 10 Jul 2016 14:22:41 -0700 (PDT)
Received: from i72vorta.tm.uni-karlsruhe.de ([141.3.71.26] helo=i72vorta.tm.kit.edu) by iramx2.ira.uni-karlsruhe.de with esmtp port 25  iface 141.3.10.81 id 1bMMBI-0004x6-Rr; Sun, 10 Jul 2016 23:22:28 +0200
Received: from [IPv6:::1] (localhost [127.0.0.1]) by i72vorta.tm.kit.edu (Postfix) with ESMTPS id 8B41EB004E5; Sun, 10 Jul 2016 23:22:28 +0200 (CEST)
To: =?UTF-8?Q?Dave_T=c3=a4ht?= <dave@taht.net>
References: <20160316000053.5163.88345.idtracker@ietfa.amsl.com> <56EEC35B.3000007@mti-systems.com> <56F0635A.7060004@student.kit.edu> <CAPRuP3k9NuTpK7WsRfOymJbTSup5aY7RgUtEDd-wsapXJxbKVg@mail.gmail.com> <1467892048445.64588@student.kit.edu> <577E4C53.9030909@uni-tuebingen.de> <cf19993f-3c00-f171-f59a-f76692d0da14@taht.net>
From: Roland Bless <roland.bless@kit.edu>
Organization: Institute of Telematics, Karlsruhe Institute of Technology (KIT)
Message-ID: <5782BC94.2040906@kit.edu>
Date: Sun, 10 Jul 2016 23:22:28 +0200
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.1) Gecko/20060111 Thunderbird/1.5 Mnenhy/0.7.3.0
MIME-Version: 1.0
In-Reply-To: <cf19993f-3c00-f171-f59a-f76692d0da14@taht.net>
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 8bit
X-ATIS-AV: ClamAV (iramx2.ira.uni-karlsruhe.de)
X-ATIS-Timestamp: iramx2.ira.uni-karlsruhe.de 1468185748.
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/psnnpOS72L_0nWRelm3akFMUBrc>
Cc: codel@lists.bufferbloat.net, "aqm@ietf.org" <aqm@ietf.org>
Subject: Re: [aqm] I-D Action: draft-ietf-aqm-codel-03.txt
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sun, 10 Jul 2016 21:22:45 -0000

Hi Dave and all,

On 07.07.2016 at 21:11 Dave Täht wrote:
> having RTOs? I do not think that a fixed delay is desirable. "You aim
> for a target", not that you need to hit it. More important is high
> utilization, and trying to keep the connections afloat, with as minimum
> latency as possible.

That is correct. There are cases where a fixed target is sub-optimal,
because one can achieve still full utilization with a lower target.
We just submitted a paper discussing several interactions
of TCP congestion control and AQMs. On the one hand results from
experiments with Codel, PIE and GSP showed that in cases with several
flows, AQMs can achieve good utilization depending on how well they
achieve loss-desynchronization. We reduced the threshold of GSP
by more than 50% and still achieved high utilization. On the
other hand, for single flows or only a few flows, AQMs cannot
achieve full utilization, because TCPs window reduction is too
large in most cases.
So it's more like having an upper bound for queueing delay
than a target and the stable point of operation may be well
below that bound.

Regards,
 Roland


From nobody Fri Jul 15 06:05:57 2016
Return-Path: <wes@mti-systems.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id CF7D812D59E for <aqm@ietfa.amsl.com>; Fri, 15 Jul 2016 06:05:55 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.601
X-Spam-Level: 
X-Spam-Status: No, score=-2.601 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, RCVD_IN_DNSWL_LOW=-0.7, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=mti-systems-com.20150623.gappssmtp.com
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 UcVCZj0T1hNZ for <aqm@ietfa.amsl.com>; Fri, 15 Jul 2016 06:05:54 -0700 (PDT)
Received: from mail-io0-x22d.google.com (mail-io0-x22d.google.com [IPv6:2607:f8b0:4001:c06::22d]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 6154A12D6A1 for <aqm@ietf.org>; Fri, 15 Jul 2016 06:05:54 -0700 (PDT)
Received: by mail-io0-x22d.google.com with SMTP id b62so103438346iod.3 for <aqm@ietf.org>; Fri, 15 Jul 2016 06:05:54 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=mti-systems-com.20150623.gappssmtp.com; s=20150623; h=to:from:subject:message-id:date:user-agent:mime-version :content-transfer-encoding; bh=/iVUzWG5tjE/15RR/fzcYVxgYcQHAJdqYAwdsgxqrKk=; b=Ypk87FlIQW2qqWwTxgYuunH6NFRpQRDxJPscsxXts7CWnflIziOewlMsTSpVuE9KN9 42M4qafZBuW+hWTnJJRk+vTZZiuc4BcE8BmKoLoiwrShq5TNHT3CRP5bb5ZqvL56L9Hj Ny6t4KfVCSzIqPatEyCoemk4RrcS/NKz/1AIXosyBXEFao29ZMuj+oEkd0f9wgdk0MHE OYgsZwKHUqt08DoICb5CCRwbRz7dOTkGOHS+BjWrUdjHaHinAjOnwB0lgS3nmZZ8X1Fb gXsGowaKXqmP7LMzRnoGCAMAxj6f309svAFLh0a2ETL6geeXF3/KmOJDU1Y1OT3lrmXD 6oTg==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:to:from:subject:message-id:date:user-agent :mime-version:content-transfer-encoding; bh=/iVUzWG5tjE/15RR/fzcYVxgYcQHAJdqYAwdsgxqrKk=; b=Q9wQY3LJcEJnmGPYGnjemdS74D5oTIR1xQqDfbADAsUHFCIo2sNK5T9z52ibKIFL8f SNzQ/StC6LkYyp/65p71MBvhPr1ZEsvHk3bLp89O8VE3EADRz4SaaCpawE/j+p79lFn/ ychCuqKTnnOXV5n2JyM+3XXmEbgJ0FpDFznq4vaoUv/Ka/OaMLks5Kx5XcKTx+N30OLl FDupbd/t8YILeLvGcPaWNAVLHFVUqBfKvE4qI7QZmw50E8x6c/AD8KsN58hnPSZO7EgA rezMd93uDgz/PONbvX1Gpct4Ohov9RvwJOfD14URjRkzpuvzmBAM3LlufHYJtNv+qY22 0m4Q==
X-Gm-Message-State: ALyK8tI5ze5cov1qEogjj5/eK8Ba/4ceqJzOAZZ3S3vtavNpx6GUj3wblrgVuN3a5jh3Zw==
X-Received: by 10.107.175.1 with SMTP id y1mr23069643ioe.137.1468587953575; Fri, 15 Jul 2016 06:05:53 -0700 (PDT)
Received: from [192.168.1.104] (cpe-76-188-215-129.neo.res.rr.com. [76.188.215.129]) by smtp.gmail.com with ESMTPSA id l195sm1421321itl.17.2016.07.15.06.05.52 for <aqm@ietf.org> (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 15 Jul 2016 06:05:52 -0700 (PDT)
To: "aqm@ietf.org" <aqm@ietf.org>
From: Wesley Eddy <wes@mti-systems.com>
Message-ID: <f6b05519-a780-43b4-74dc-4b6c883e3639@mti-systems.com>
Date: Fri, 15 Jul 2016 09:05:48 -0400
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:45.0) Gecko/20100101 Thunderbird/45.2.0
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/83GZuu7QTG78Nf0FQU4IgLmj6iU>
Subject: [aqm] meeting update
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 15 Jul 2016 13:05:56 -0000

There's an update pending to the agenda for the AQM meeting.

It is currently planned for AQM to use the last 45 minutes of the TSVWG 
timeslot on Friday morning.  The ADs have send a request to update the 
agenda, and the TSVWG agenda includes it already: 
https://datatracker.ietf.org/meeting/96/agenda/tsvwg/

There didn't appear to be any top-order scheduling conflicts with this 
move, and it was assumed that most of the people interested in the AQM 
meeting were planning to be in TSVWG as well. Apologies if this was a 
bad assumption.

Sorry for any confusion this will cause!  We'll try to make sure it's 
marked clearly on the agenda site.


From nobody Thu Jul 28 10:27:07 2016
Return-Path: <dave.taht@gmail.com>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D5F9D12D51B; Thu, 28 Jul 2016 10:27:05 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.7
X-Spam-Level: 
X-Spam-Status: No, score=-2.7 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, FREEMAIL_FROM=0.001, RCVD_IN_DNSWL_LOW=-0.7, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com
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 isI4ZQKkff7B; Thu, 28 Jul 2016 10:27:04 -0700 (PDT)
Received: from mail-io0-x243.google.com (mail-io0-x243.google.com [IPv6:2607:f8b0:4001:c06::243]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 159BA12D89C; Thu, 28 Jul 2016 10:27:04 -0700 (PDT)
Received: by mail-io0-x243.google.com with SMTP id y34so7976167ioi.3; Thu, 28 Jul 2016 10:27:04 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113;  h=mime-version:from:date:message-id:subject:to :content-transfer-encoding; bh=9lUcIwZsX06uzY42iIBxBm5+bEUeAb+2WJVbcc0rXJE=; b=quCxpvp2MWck7p8TM77SAV3G+ZWjBLUN9Dt7M6wThONL8iG2Zf2mB4o+qmvJfPr4UQ pL+Itax9Zn9ZB/dWdeYRQMOL0OUtKytMK2ezXedzVrFU1OMzPFg5eDlovNeeYeH3Fo5Q qvrD2oqTUPLaFvD/YNPJWE2FpUOPxlx+mAS/SwH10lVMDwr6AylWbgrVyEOVlQVhPKog hwMPLy9KuvHnBuAwxjFSz8tVlDH+50G1uqccHKzajD6XE1dGYR+ijsOie1H7NCbwWTtZ 2TGnQlCwCq3zcyHcSHtX6RMvYxnHKvgi7e16yV52vh9c7tcwMNLEOF8M9V0PZ86B0Mtb ykwQ==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:from:date:message-id:subject:to :content-transfer-encoding; bh=9lUcIwZsX06uzY42iIBxBm5+bEUeAb+2WJVbcc0rXJE=; b=C7IoN5UDUzQ3F+axH2KRM+84w05/dPvFsJ7meTbqfMVCnkQC94JE+Vq3v/jo0Xi/Fv s9NX9BOoNLza+Ofg2BoWur+bNAyK74bhjWcW7Ue724nU+dhrJO7Povpbk2OpTr+y1Fsy GYYbanC6ZgDQl5FAGYzSotZsOYRzK4rdceX6FiTA7PR5+//6yJGzOENtuaUwZEvkRN0T zQl860PmNfRNin6PJqUR/WmYuX222Y2McZ1JkqI9A7JlvAJRgGhKl2PTf9iMrrj0HZel i4I2r/BFXht0O2d+4px+gd/GP9nRUjQA4/xdiXl7t+5n30RdfJL2s13xx/c5kXwR4A2E FAow==
X-Gm-Message-State: AEkoouuDugR0DM3vkhB68SExezySXzCtai3ctUcgwBM7pTHtJNJahR41elI8/QUswDljFOG3fLklit+qVa0Yfw==
X-Received: by 10.107.2.78 with SMTP id 75mr39173863ioc.128.1469726823231; Thu, 28 Jul 2016 10:27:03 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.79.14.17 with HTTP; Thu, 28 Jul 2016 10:27:02 -0700 (PDT)
From: Dave Taht <dave.taht@gmail.com>
Date: Thu, 28 Jul 2016 19:27:02 +0200
Message-ID: <CAA93jw48xb7g-q839iixUadRG9GH9e_W5KvjPtpKKij=UmOJvg@mail.gmail.com>
To: "aqm@ietf.org" <aqm@ietf.org>, tcpprague@ietf.org, bloat <bloat@lists.bufferbloat.net>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/qclC8HP3ShW12F8Qo6K8Fr7dBxQ>
Subject: [aqm] what is the correct linux tc u32 match to ignore ecn but preserve tos?
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Thu, 28 Jul 2016 17:27:06 -0000

I can't ever get my endianess straight, and I figure that someone here
might "just know"

Do I use match u8 0xfc or 0xcf?

tc filter add dev $DEV parent 1:0 protocol ip prio 10 u32 \
      match ip tos 0x10 0xff  flowid 1:10

(am trying to revise this from in front of a mac:
https://www.bufferbloat.net/projects/cerowrt/wiki/Wondershaper_Must_Die/
)

Bonus points if you can come up with the right tc filter line for ipv6.

What I would probably do is have a filter chain to then classify the
thing further for L4s, but I don't understand how or where I'd direct
CE markings correctly. (?)

How does it happen in the pi2 code? is that on github somewhere?

--=20
Dave T=C3=A4ht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org=E3=AF=9D=E3=AF=9F


From nobody Fri Jul 29 15:11:06 2016
Return-Path: <toke@toke.dk>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id D827B12D8D0; Fri, 29 Jul 2016 15:11:00 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.002
X-Spam-Level: 
X-Spam-Status: No, score=-2.002 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001] autolearn=ham autolearn_force=no
Authentication-Results: ietfa.amsl.com (amavisd-new); dkim=pass (1024-bit key) header.d=toke.dk
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 4ffDO4xmKZCO; Fri, 29 Jul 2016 15:10:58 -0700 (PDT)
Received: from mail2.tohojo.dk (mail2.tohojo.dk [77.235.48.147]) (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 7391812B056; Fri, 29 Jul 2016 15:10:58 -0700 (PDT)
X-Virus-Scanned: amavisd-new at mail2.tohojo.dk
DKIM-Filter: OpenDKIM Filter v2.10.3 mail2.tohojo.dk D273841622
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=toke.dk; s=201310; t=1469830254; bh=1mHERhiga9KJEVbt6X2AVPGkaiZkes8UqsSn16pC4uA=; h=From:To:Subject:Date:From; b=smLQ6P4rL1NCP2xBFMQa4kCzVzMoNMhMbf1z9sk+Mw3cU5A2+g2SEkpIEYuLrBh32 9Lsdi2wFuUE9gCiJZEIfnJfDAMLLF9i0q+wax5oXgWBE/zF1WcYwG03G/8TyB65571 bZNZ+nGkVAL5gGd0cmImtsMpwByy4rcl8FiX6pgk=
Sender: toke@toke.dk
Received: by alrua-x1.borgediget.toke.dk (Postfix, from userid 1000) id CDFA13F47; Sat, 30 Jul 2016 00:10:53 +0200 (CEST)
From: =?utf-8?Q?Toke_H=C3=B8iland-J=C3=B8rgensen?= <toke@toke.dk>
To: tsvwg@ietf.org, aqm@ietf.org
Date: Sat, 30 Jul 2016 00:10:53 +0200
X-Clacks-Overhead: GNU Terry Pratchett
Message-ID: <877fc4jdsi.fsf@toke.dk>
MIME-Version: 1.0
Content-Type: text/plain
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/_N-yFYE9YKwBcmJuOU4GOPPMWks>
Subject: [aqm] Applying AQM to the WiFi MAC layer
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Fri, 29 Jul 2016 22:11:01 -0000

Since Andrew McGregor mentioned this at the mic during the tsvwg/aqm
session in Berlin, I thought I'd post some references to the work on
fixing queueing behaviour on WiFi (by incorporating principles from
FQ-CoDel in the Linux WiFi stack at the MAC layer).

There are still some bugs to work out, but these two blog entries
provide some details on the progress so far:

https://blog.tohojo.dk/2016/06/fixing-the-wifi-performance-anomaly-on-ath9k.html

http://blog.cerowrt.org/post/fq_codel_on_ath10k/

Cheers,

-Toke


From nobody Fri Jul 29 22:17:25 2016
Return-Path: <wwwrun@rfc-editor.org>
X-Original-To: aqm@ietfa.amsl.com
Delivered-To: aqm@ietfa.amsl.com
Received: from localhost (localhost [127.0.0.1]) by ietfa.amsl.com (Postfix) with ESMTP id E363A12D5AF; Fri, 29 Jul 2016 22:17:18 -0700 (PDT)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -103.889
X-Spam-Level: 
X-Spam-Status: No, score=-103.889 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RCVD_IN_DNSWL_LOW=-0.7, RP_MATCHES_RCVD=-1.287, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001, USER_IN_WHITELIST=-100] autolearn=ham autolearn_force=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 MdsWX8eBHX0P; Fri, 29 Jul 2016 22:17:12 -0700 (PDT)
Received: from rfc-editor.org (rfc-editor.org [4.31.198.49]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by ietfa.amsl.com (Postfix) with ESMTPS id 7954C12D61D; Fri, 29 Jul 2016 22:16:59 -0700 (PDT)
Received: by rfc-editor.org (Postfix, from userid 30) id 5DA70B8032A; Fri, 29 Jul 2016 22:16:59 -0700 (PDT)
To: ietf-announce@ietf.org, rfc-dist@rfc-editor.org
X-PHP-Originating-Script: 1005:ams_util_lib.php
From: rfc-editor@rfc-editor.org
Message-Id: <20160730051659.5DA70B8032A@rfc-editor.org>
Date: Fri, 29 Jul 2016 22:16:59 -0700 (PDT)
Archived-At: <https://mailarchive.ietf.org/arch/msg/aqm/MZk6djvTF1i3O-s3veHSX8tRMqU>
Cc: drafts-update-ref@iana.org, aqm@ietf.org, rfc-editor@rfc-editor.org
Subject: [aqm] RFC 7928 on Characterization Guidelines for Active Queue Management (AQM)
X-BeenThere: aqm@ietf.org
X-Mailman-Version: 2.1.17
Precedence: list
List-Id: "Discussion list for active queue management and flow isolation." <aqm.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/options/aqm>, <mailto:aqm-request@ietf.org?subject=unsubscribe>
List-Archive: <https://mailarchive.ietf.org/arch/browse/aqm/>
List-Post: <mailto:aqm@ietf.org>
List-Help: <mailto:aqm-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/aqm>, <mailto:aqm-request@ietf.org?subject=subscribe>
X-List-Received-Date: Sat, 30 Jul 2016 05:17:19 -0000

A new Request for Comments is now available in online RFC libraries.

        
        RFC 7928

        Title:      Characterization Guidelines for Active Queue 
                    Management (AQM) 
        Author:     N. Kuhn, Ed.,
                    P. Natarajan, Ed.,
                    N. Khademi, Ed.,
                    D. Ros
        Status:     Informational
        Stream:     IETF
        Date:       July 2016
        Mailbox:    nicolas.kuhn@cnes.fr, 
                    prenatar@cisco.com, 
                    naeemk@ifi.uio.no,
                    dros@simula.no
        Pages:      37
        Characters: 90255
        Updates/Obsoletes/SeeAlso:   None

        I-D Tag:    draft-ietf-aqm-eval-guidelines-13.txt

        URL:        https://www.rfc-editor.org/info/rfc7928

        DOI:        http://dx.doi.org/10.17487/RFC7928

Unmanaged large buffers in today's networks have given rise to a slew
of performance issues.  These performance issues can be addressed by
some form of Active Queue Management (AQM) mechanism, optionally in
combination with a packet-scheduling scheme such as fair queuing.
This document describes various criteria for performing
characterizations of AQM schemes that can be used in lab testing
during development, prior to deployment.

This document is a product of the Active Queue Management and Packet Scheduling Working Group of the IETF.


INFORMATIONAL: This memo provides information for the Internet community.
It does not specify an Internet standard of any kind. Distribution of
this memo is unlimited.

This announcement is sent to the IETF-Announce and rfc-dist lists.
To subscribe or unsubscribe, see
  https://www.ietf.org/mailman/listinfo/ietf-announce
  https://mailman.rfc-editor.org/mailman/listinfo/rfc-dist

For searching the RFC series, see https://www.rfc-editor.org/search
For downloading RFCs, see https://www.rfc-editor.org/retrieve/bulk

Requests for special distribution should be addressed to either the
author of the RFC in question, or to rfc-editor@rfc-editor.org.  Unless
specifically noted otherwise on the RFC itself, all RFCs are for
unlimited distribution.


The RFC Editor Team
Association Management Solutions, LLC


