Skip to main content

Dual-Queue Coupled Active Queue Management (AQM) for Low Latency, Low Loss, and Scalable Throughput (L4S)
draft-ietf-tsvwg-aqm-dualq-coupled-25

Yes

(Martin Duke)

No Objection

Erik Kline
John Scudder
(Alvaro Retana)
(Andrew Alston)

Note: This ballot was opened for revision 24 and is now closed.

Zaheduzzaman Sarker
Yes
Comment (2022-08-25 for -24) Not sent
Thanks for working on this specification and for being persuasive.
Erik Kline
No Objection
John Scudder
No Objection
Paul Wouters
No Objection
Comment (2022-08-24 for -24) Sent
# Security AD review of draft-ietf-tsvwg-aqm-dualq-coupled-24

CC @paulwouters

The queuing details are definitely outside my area of expertise. It seems reasonable but I would not add too much weight to this opinion.

## COMMENTS

### Abstract

The abstract is very long. I would cut it at "The coupling acts like a"

### Section 1 repetition

There is a lot of repeated text from the architecture document, eg
about speed of flows and 360 racing cameras that could probably
be omitted in this document and replaced with a "the reader should
be familiar with [L4S architecture doc]. That would cut out a lot
of Section 1 and make the document more readable to those readers
who (like me) just had to read the architecture document as well.
Roman Danyliw
No Objection
Comment (2022-08-24 for -24) Sent
** Section 1.1
   In the developed world, further
   increases in access network bit-rate offer diminishing returns,
   whereas latency is still a multi-faceted problem.   

Is there a reference for this assertion?  Generalizing to the “developed world” seems difficult.

** Section 4.1
   It also
   explains that, on the current Internet, scheduling usually enforces
   separation between 'sites' (e.g. households, businesses or mobile
   users), but it is not common to need to schedule or police individual
   application flows.

I have no data on this, but is this suggesting that carriers don’t have agreements to carry certain classes of traffic with more priority on a given pipe?

** Section 4.1.  

   By the above arguments, per-flow policing might not be necessary and
   in trusted environments  it is certainly unlikely to be needed.

Could the properties of a “trusted environment” be described in a bit more detail?

Nits
** Section 1.2. Please provide a reference to BBRv2.

** Section 1.4.  s/watchings/watching/
Martin Duke Former IESG member
Yes
Yes (for -24) Unknown

                            
Alvaro Retana Former IESG member
No Objection
No Objection (for -24) Not sent

                            
Andrew Alston Former IESG member
No Objection
No Objection (for -24) Not sent

                            
Lars Eggert Former IESG member
No Objection
No Objection (2022-08-24 for -24) Sent
# GEN AD review of draft-ietf-tsvwg-aqm-dualq-coupled-24

CC @larseggert

Thanks to Christer Holmberg for the General Area Review Team (Gen-ART) review
(https://mailarchive.ietf.org/arch/msg/gen-art/1WR0V1l2Vh7yUfnVUwzS1t3H2MY).

## Comments

### Boilerplate

This document uses the RFC2119 keywords ['MUST', 'SHALL', 'SHALL NOT', 'SHOULD
NOT', 'RECOMMENDED', 'OPTIONAL', 'MAY', 'MUST NOT', 'REQUIRED', 'SHOULD'], but
does not contain the recommended RFC8174 boilerplate. (It contains some text
with a similar beginning.)

### Inclusive language

Found terminology that should be reviewed for inclusivity; see
https://www.rfc-editor.org/part2/#inclusive_language for background and more
guidance:

 * Term `traditionally`; alternatives might be `classic`, `classical`,
   `common`, `conventional`, `customary`, `fixed`, `habitual`, `historic`,
   `long-established`, `popular`, `prescribed`, `regular`, `rooted`,
   `time-honored`, `universal`, `widely used`, `widespread`
 * Term `native`; alternatives might be `built-in`, `fundamental`, `ingrained`,
   `intrinsic`, `original`

## Nits

All comments below are about very minor potential issues that you may choose to
address in some way - or ignore - as you see fit. Some were flagged by
automated tools (via https://github.com/larseggert/ietf-reviewtool), so there
will likely be some false positives. There is no need to let me know what you
did with these suggestions.

### Outdated references

Reference `[RFC2309]` to `RFC2309`, which was obsoleted by `RFC7567` (this may
be on purpose).

### URLs

Found non-HTTP URLs in the document:

 * speedtest.net

These URLs in the document did not return content:

 * https://www.bell-labs.com/usr/koen.de_schepper
 * http://www.icir.org/floyd/red.html
 * http://www.hpcc.jp/pfldnet2009/Program_files/1569198525.pdf

These URLs in the document can probably be converted to HTTPS:

 * http://dl.acm.org/citation.cfm?doid=2910017.2910633
 * http://queue.acm.org/issuedetail.cfm?issue=2208917
 * http://infocom2003.ieee-infocom.org/papers/27_04.PDF
 * http://bobbriscoe.net/

### Grammar/style

#### Section 1.1, paragraph 4
```
euing delay of about 20ms on average but there are also regular 25ms delay s
                                    ^^^^
```
Use a comma before "but" if it connects two independent clauses (unless they
are closely connected and short).

#### Section 1.1, paragraph 4
```
ems. DCTCP teaches us that two small but radical changes to congestion contro
                                    ^^^^
```
Use a comma before "but" if it connects two independent clauses (unless they
are closely connected and short).

#### Section 1.2, paragraph 5
```
equires little extra processing. Therefore it is believed the Coupled AQM wo
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 1.3, paragraph 2
```
svwg-l4s-arch] and in [RFC3649]. Therefore control of queuing and utilizatio
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 1.3, paragraph 6
```
e.g. DNS, VoIP, game sync datagrams, etc). The DualQ Coupled AQM behaviour i
                                     ^^^
```
A period is needed after the abbreviation "etc.".

#### Section 1.3, paragraph 10
```
me throughput whichever it uses. Therefore both queues can feed into the ful
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 1.3, paragraph 13
```
 on the path, for instance by legacy WiFi equipment. However, if L4S support 
                                     ^^^^
```
Did you mean "Wi-Fi"? (This is the officially approved term by the Wi-Fi
Alliance.).

#### Section 1.3, paragraph 14
```
uce the burstiness of the _incoming_ WiFi. Also, trials of WiFi equipment wit
                                     ^^^^
```
Did you mean "Wi-Fi"? (This is the officially approved term by the Wi-Fi
Alliance.).

#### Section 1.4, paragraph 1
```
he _incoming_ WiFi. Also, trials of WiFi equipment with an L4S DualQ Coupled 
                                    ^^^^
```
Did you mean "Wi-Fi"? (This is the officially approved term by the Wi-Fi
Alliance.).

#### Section 1.4, paragraph 1
```
DualQ Coupled AQM on the _outgoing_ WiFi interface are in progress, and earl
                                    ^^^^
```
Did you mean "Wi-Fi"? (This is the officially approved term by the Wi-Fi
Alliance.).

#### Section 1.4, paragraph 3
```
stick to the user's finger on the touch pad and the experience fed from the r
                                  ^^^^^^^^^
```
This word is normally spelled as one.

#### Section 1.4, paragraph 3
```
ves low latency to L4S traffic whether or not there is Classic traffic. The 
                               ^^^^^^^^^^^^^^
```
Consider shortening this phrase to just "whether". It is correct though if you
mean "regardless of whether".

#### Section 2.1, paragraph 3
```
ue to prevent under-utilization. Therefore a separate queue is provided for 
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 2.4, paragraph 1
```
rienced by L4S and Classic traffic. Thus k indirectly determines the ratio be
                                    ^^^^
```
A comma may be missing after the conjunctive/linking adverb "Thus".

#### Section 2.4, paragraph 9
```
ffic is controlled by the coupling. Also there only has to be a coupling in o
                                    ^^^^
```
A comma may be missing after the conjunctive/linking adverb "Also".

#### Section 2.4, paragraph 14
```
nsitive to link rate and it requires less operations per packet than RED. How
                                     ^^^^
```
Did you mean "fewer"? The noun operations is countable.

#### Section 2.5.1.1, paragraph 11
```
n a simple accumulating counter. Therefore the type of queue delay statistic
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 2.5.2.1, paragraph 3
```
elay. This could be done with a small number of bins with configurable edges 
                              ^^^^^^^^^^^^^^^^^
```
Specify a number, remove phrase, use "a few", or use "some".

#### Section 2.5.2.1, paragraph 8
```
ion in overload state is accumulated but no new report is generated until th
                                    ^^^^
```
Use a comma before "but" if it connects two independent clauses (unless they
are closely connected and short).

#### Section 2.5.2.1, paragraph 10
```
s management requirements. The raison d'etre of the DualQ Coupled AQM is to 
                               ^^^^^^^^^^^^^
```
"raison d'etre" is an imported foreign name or expression, which originally has
a diacritic.

#### Section 2.5.2.1, paragraph 11
```
ale with bandwidth-delay product. Therefore there is no need to repeat these 
                                  ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### Section 2.5.2.2, paragraph 6
```
avoid complexity and unintended side-effects with per-flow policing, it need
                                ^^^^^^^^^^^^
```
Did you mean "side effects" (=adverse effect, unintended consequence)? Open
compounds are not hyphenated.

#### Section 4.1, paragraph 4
```
g dequeued. There is then the question of whether to sacrifice L4S throughpu
                          ^^^^^^^^^^^^^^^^^^^^^^^
```
Wordiness: Consider shortening this phrase.

#### Section 4.1, paragraph 4
```
S throughput: By using weighted round robin as the conditional priority sche
                                ^^^^^^^^^^^
```
This word is normally spelled with a hyphen.

#### Section 4.2.2, paragraph 3
```
s-id]). Saturation raises the question of whether to relieve congestion by i
                          ^^^^^^^^^^^^^^^^^^^^^^^
```
Wordiness: Consider shortening this phrase.

#### Section 4.2.2, paragraph 7
```
S or Classic. This raises the question of whether and when to introduce drop
                          ^^^^^^^^^^^^^^^^^^^^^^^
```
Wordiness: Consider shortening this phrase.

#### Section 4.2.3.1, paragraph 1
```
TCP BBR v2 Alpha/Preview Release", github repository; Linux congestion contr
                                   ^^^^^^
```
The official name of this software platform is spelled with a capital "H".

#### Section 5, paragraph 2
```
 Coupled Active Queue Management", Masters Thesis, Dept of Informatics, Uni O
                                   ^^^^^^^
```
For an academic degree, use "Master's".

#### Section 5, paragraph 2
```
st, P. and J. Morton, "L4S Tests", github README, August 2021, <https://githu
                                   ^^^^^^
```
The official name of this software platform is spelled with a capital "H".

#### Section 7.2, paragraph 18
```
[SCReAM] Johansson, I., "SCReAM", github repository; , <https://github.com/Er
                                  ^^^^^^
```
The official name of this software platform is spelled with a capital "H".

#### "Appendix A.", paragraph 1
```
ts, 1 MTU of space is always allowed and the limit is deliberately tested be
                                    ^^^^
```
Use a comma before "and" if it connects two independent clauses (unless they
are closely connected and short).

#### "A.1.", paragraph 22
```
on controllers conducted by Mishra _et al_ in Jul-Oct 2019 [CCcensus19], most
                                    ^^^^^
```
A period is misplaced or missing.

#### "A.1.", paragraph 24
```
tion of RTT_typ around its mean. Otherwise the target queue would only avoid
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Otherwise".

#### "A.1.", paragraph 24
```
ween a user and their local CDN. Currently no data is available on the varia
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Currently".

#### "A.1.", paragraph 25
```
 g * f = 25ms * 0.38 * 2 = 19 ms. However a further adjustment is warranted,
                                  ^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "However".

#### "A.1.", paragraph 25
```
ranted, because target is moving year on year. The paper is based on data col
                                 ^^^^^^^^^^^^
```
This word is normally spelled with a hyphen.

#### "A.1.", paragraph 25
```
 (mobile) between 2020 and 2021. Therefore we recommend a default of target 
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Therefore".

#### "A.1.", paragraph 34
```
this value. The propagation delay half way round the planet and back in glass
                                  ^^^^^^^^
```
The adjective or adverb "halfway" is spelled as one word.

#### "A.1.", paragraph 45
```
hen the packet is dequeued. If time- stamping is not easy to introduce with c
                               ^^^^^^^^^^^^^^
```
This word seems to be formatted incorrectly. Consider fixing the spacing or
removing the hyphen completely.

#### "A.1.", paragraph 48
```
available space. The choice is a trade off; a shared buffer can use less mem
                                 ^^^^^^^^^
```
When "trade-off" is used as a noun or modifier, it needs to be hyphenated.

#### "A.1.", paragraph 50
```
etails of both edge-cases added. Similarly Figure 8 repeats the core PI algo
                                 ^^^^^^^^^
```
A comma may be missing after the conjunctive/linking adverb "Similarly".

#### "A.1.", paragraph 66
```
ghted scheduler such as weighted round robin (WRR) is recommended. As long a
                                 ^^^^^^^^^^^
```
This word is normally spelled with a hyphen.

#### "B.1.", paragraph 27
```
k node cannot easily know the RTT of any of the flows anyway. RTT-dependence
                                  ^^^^^^^^^
```
Consider simply using "of" instead.

#### "B.2.", paragraph 3
```
because we only use one point - at the the typical base RTT where the operato
                                   ^^^^^^^
```
Possible typo: you repeated a word.

## Notes

This review is in the ["IETF Comments" Markdown format][ICMF], You can use the
[`ietf-comments` tool][ICT] to automatically convert this review into
individual GitHub issues. Review generated by the [`ietf-reviewtool`][IRT].

[ICMF]: https://github.com/mnot/ietf-comments/blob/main/format.md
[ICT]: https://github.com/mnot/ietf-comments
[IRT]: https://github.com/larseggert/ietf-reviewtool