TCP Alternative Backoff with ECN (ABE)
draft-ietf-tcpm-alternativebackoff-ecn-12

[Ballot comment]
Update: Adam pointed out I thanked the authors for an empty IANA section. I meant section _6_, the Experiment Goals section.

Thank you for section 8.

Please expand PIE and CoDel on first mention.

[Ballot comment]
Thanks for a well-written and easy-to-understand document. I find the proposed
experiment quite interesting.

I have a small number of editorial nits that you may want to address if you
produce another version of the document.

---------------------------------------------------------------------------

I-D Nits reports:

  == Unused Reference: 'RFC7713' is defined on line 538, but no explicit
    reference was found in the text

---------------------------------------------------------------------------

General:

  When used as a compound adjective, "congestion control" should be hyphenated.
  e.g.:

  * congestion-control algorithm(s)
  * congestion-control mechanism(s)
  * congestion-control response
  * congestion-control behaviour

---------------------------------------------------------------------------

§1:

>  Recognizing these changes in modern AQM practices,
>  the strict requirement that ECN CE signals be treated identically to
>  inferred packet loss have been relaxed [RFC8311].  This document

Nit: "...has been..."

---------------------------------------------------------------------------

§12.2:

>  [BUFFERBLOAT]
>            Gettys, J. and K. Nichols, "Bufferbloat: Dark Buffers in
>            the Internet", November 2011.

This seems incomplete. I would suggest something more like:

  [BUFFERBLOAT]
              Gettys, J. and K. Nichols, "Bufferbloat: Dark Buffers in the
              Internet", ACM Queue 9, 11, DOI 10.1145/2063166.2071893,
              November 2011, https://queue.acm.org/detail.cfm?id=2071893

[Ballot comment]
This looks good and I'm looking forward to seeing the report on the experiment's results.

I just had one place where I was perhaps confused by the text, in Section 1:

  effectively doubles the amount of data that can be in flight, the
  maximum round-trip time (RTT) experience, and the path's effective
  RTT using the network path.

I'm probably just misreading, but "the path's effective RTT using the network path"
sounds odd with the two "paths" in there.  Is the last one referring to the
bottleneck part?

(Via drafts-lastcall@iana.org): IESG/Authors/WG Chairs:

The IANA Functions Operator has reviewed draft-ietf-tcpm-alternativebackoff-ecn-09, which is currently in Last Call, and has the following comments:

We understand that this document doesn't require any registry actions.

While it's often helpful for a document's IANA Considerations section to remain in place upon publication even if there are no actions, if the authors strongly prefer to remove it, we do not object.

If this assessment is not accurate, please respond as soon as possible.

Thank you,

Sabrina Tanamal
Senior IANA Services Specialist

The following Last Call announcement was sent out (ends 2018-08-28):

From: The IESG
To: IETF-Announce
CC: tcpm@ietf.org, tuexen@fh-muenster.de, draft-ietf-tcpm-alternativebackoff-ecn@ietf.org, ietf@kuehlewind.net, =?utf-8?q?Michael_T=C3=BCxen?= , tcpm-chairs@ietf.org
Reply-To: ietf@ietf.org
Sender:
Subject: Last Call:  (TCP Alternative Backoff with ECN (ABE)) to Experimental RFC


The IESG has received a request from the TCP Maintenance and Minor Extensions
WG (tcpm) to consider the following document: - 'TCP Alternative Backoff with
ECN (ABE)'
  as Experimental RFC

The IESG plans to make a decision in the next few weeks, and solicits final
comments on this action. Please send substantive comments to the
ietf@ietf.org mailing lists by 2018-08-28. Exceptionally, comments may be
sent to iesg@ietf.org instead. In either case, please retain the beginning of
the Subject line to allow automated sorting.

Abstract


  Active Queue Management (AQM) mechanisms allow for burst tolerance
  while enforcing short queues to minimise the time that packets spend
  enqueued at a bottleneck.  This can cause noticeable performance
  degradation for TCP connections traversing such a bottleneck,
  especially if there are only a few flows or their bandwidth-delay-
  product is large.  The reception of a Congestion Experienced (CE) ECN
  mark indicates that an AQM mechanism is used at the bottleneck, and
  therefore the bottleneck network queue is likely to be short.
  Feedback of this signal allows the TCP sender-side ECN reaction in
  congestion avoidance to reduce the Congestion Window (cwnd) by a
  smaller amount than the congestion control algorithm's reaction to
  inferred packet loss.  This specification therefore defines an
  experimental change to the TCP reaction specified in RFC3168, as
  permitted by RFC 8311.




The file can be obtained via
https://datatracker.ietf.org/doc/draft-ietf-tcpm-alternativebackoff-ecn/

IESG discussion can be tracked via
https://datatracker.ietf.org/doc/draft-ietf-tcpm-alternativebackoff-ecn/ballot/


No IPR declarations have been submitted directly on this I-D.

1. Summary

The document shepherd is Michael Tüxen .

The responsible Area Director is Mirja Kühlewind .

The recently published RFC 8311 (Relaxing Restrictions on Explicit Congestion Notification (ECN) Experimentation) allows a TCP sender to react to an ECN signal in a different way than to a inferred loss signal if the behaviour is documented in an experimental RFC. This document describes such an experiment in which the reduction of the congestion window in response to an ECN signal is smaller than in response to an inferred packet loss. Therefore the TCPM working group requests publication as an experimental document.

2. Review and Consensus

The document has been discussed several times in working group meetings without any major controversy. There were also no substantial problems reported during the working group last call and it is implemented for FreeBSD (the code has been committed to the FreeBSD code base). The risk of running this experiment is relatively low, since the reaction to a loss is not changed. There is very strong consensus in the TCPM working group that this document should be published.

3. Intellectual Property

Each author has stated that their direct, personal knowledge of any IPR related to this document has already been disclosed, in conformance with BCPs 78 and 79.
None of the authors is aware of any IPR related to this document.

4. Other Points

None.