Early Review of draft-ietf-idr-bgp-flowspec-oid-11
review-ietf-idr-bgp-flowspec-oid-11-rtgdir-early-huston-2020-07-06-00

(copy to rtg-dir - any followups should include as cc idr-chairs@ietf.org,
draft-djsmith-bgp-flowspec-oid.all@ietf.org)

Hello

I have been selected to do a routing directorate “early” review of this draft:
https://datatracker.ietf.org/doc/draft-ietf-idr-bgp-flowspec-oid

The routing directorate will, on request from the working group chair, perform
an “early” review of a draft before it is submitted for publication to the
IESG. The early review can be performed at any time during the draft’s lifetime
as a working group document. The purpose of the early review depends on the
stage that the document has reached.

A co-chair of the IDR, Susan Hares, has requested an early review of this
document

For more information about the Routing Directorate, please see
http://trac.tools.ietf.org/area/rtg/trac/wiki/RtgDir

Document: draft-ietf-idr-bgp-flowspec-oid-11.txt
Reviewer: Geoff Huston
Review Date: 3 July 2020
Intended Status: Standards Track

Summary:

I have some major concerns about this document, mainly relating to updates to
an RFC-to-be even before the original document has been published as an RFC. I
assume that there are reasons why the original document
(draft-ietf-idr-rfc5575bis) was not updated directly, particularly given that
the initial drafts of this validation revision document predates the initial
drafts rfc5575bis document. It is therefore unclear why the process of review
of the 5575bis document has not adopted this revised specification during the
course of the development of the bis document during its development. The
question I am left with is: What has changed in recent times that would make
revision of a published RFC more likely than revision of an in-progress working
group draft?

Comments:

The document clearly describes the motivation of the revised validation 
procedures (reduced operational complexity for dissemination of BGP policy
rules when using a route controller).

The proposed revision is susceptible to misunderstanding. It takes the set of
three conditions specified in RFC5575bis, where all the conditions must hold
for the Flow Specification NLRI to be validated, and altering one of these
condition to be two conditions, either of which can hold. The three further
comments (labelled 1 through 3 in section 4.1) are unclear in their intent. It
is also unclear why the second paragraph in section 4.1 is labelled "a." when
there is no subsequent section labelled "b.". The indentation of the succeeding
paras of section 4.1 suggest that they are part of the redefinition of step(b)
of the validation procedures in RFC5575bis, but the text is descriptive rather
than prescriptive and the cumulative intent of this proposed revision is
entirely unclear to this reviewer.

It is unfortunate that RFC5575bis chose to add further validation conditions
beside the three conditions labelled (a), (b) and (c) in the body of the text
in Section 6. This refinement to the original three conditions (a.,b., and c.)
is unclear. The revised text in this draft makes this state even more
confusing. If the intent of this draft was to clarify the intent of RFC5575 as
well as adding additional criteria for validation, then this does not appear to
have been achieved.

It is unclear for me that the publication of RFC5575bis and the publication of
this draft as further revision to RFC5575bis serves the purpose of extending
the utility of the Flowspec validation procedure in a useful manner. If 5575bis
is flawed in its description of a clear unambiguous validation procedure then
its process through to publication should be halted and the document passed
back to the WG. If, on the other hand, the WG is happy with the bis document in
its current states, and happy in the light of the knowledge of the existence of
this revisionm draft, then it would appear that the working group is unwilling
to contemplate the revisions proposed in this draft, and the intent of this
draft as a working gropup document is at best confusing.