On Mar 22, 2025, at 4:56 AM, Evans, John <jevanamz=40amazon.co...@dmarc.ietf.org
<mailto:40amazon.co...@dmarc.ietf.org>> wrote:
Hi All,
Based on feedback from the last meeting we moved from option 1 to option 2
below. Rob Wilton proposed option 3 in the discussion at the WG meeting this
week. Hence, we would like to solicit feedback from the group on whether there
is consensus with the current approach (option 2 below), or whether there is
support to transition to option 3.
As an author, I think there's value both in separating concerns and in making
incremental progress, hence my preference is to stick with option 2,
1. Previous approach: separate drafts for each IM and DM, e.g.
a. draft-ietf-opsawg-discardmodel
i. IM (YANG)
b. draft-ietf-opsawg-…
i. IM (references draft-ietf-opsawg-discardmodel)
ii. DM (YANG)
c. draft-evans-opsawg-ipfix-discard-class-ie
i. IM (references draft-ietf-opsawg-discardmodel)
ii. DM (IPFIX IE)
2. Current approach: one draft for YANG IM and YANG DM with other DMs in
subsequent drafts:
a. draft-ietf-opsawg-discardmodel:
i. IM (YANG)
ii. DM (YANG)
b. draft-evans-opsawg-ipfix-discard-class-ie
i. IM (references draft-ietf-opsawg-discardmodel)
ii. DM (IPFIX IE)
3. Alternative approach proposed by Rob Wilton: single draft for IMs and DMs,
e.g.
a. draft-ietf-opsawg-discardmodel
i. IM (YANG)
ii. DM (YANG)
iii. DM (IPFIX IE)
cheers
John
-----------------------
From: "Evans, John" <jevan...@amazon.co.uk <mailto:jevan...@amazon.co.uk>>
Date: Tuesday 4 March 2025 at 10:29
To: "mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>" <mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>>, Benoit Claise
<benoit.cla...@huawei.com <mailto:benoit.cla...@huawei.com>>, "opsawg@ietf.org <mailto:opsawg@ietf.org>" <opsawg@ietf.org <mailto:opsawg@ietf.org>>,
"opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>" <opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>>, Jeff Haas <jh...@juniper.net
<mailto:jh...@juniper.net>>, "Pylypenko, Oleksandr" <o...@amazon.com <mailto:o...@amazon.com>>
Cc: "draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>" <draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>>
Subject: Re: draft-ietf-opsawg-discardmodel
Hi All,
We've checked in a updated version of draft-ietf-opsawg-discardmodel. The most
significant change is that it incorporates both the YANG information model, and the
YANG data model derived from the information model – as was proposed after the last
meeting:https://datatracker.ietf.org/doc/draft-ietf-opsawg-discardmodel/
<https://datatracker.ietf.org/doc/draft-ietf-opsawg-discardmodel/>
We have also checked in a complementary draft, which defines a new IPFIX Information
Element for classifying flow-level discards that aligns with the information model
defined in
draft-ietf-opsawg-discardmodel:https://datatracker.ietf.org/doc/draft-evans-opsawg-ipfix-discard-class-ie/
<https://datatracker.ietf.org/doc/draft-evans-opsawg-ipfix-discard-class-ie/>
We would appreciate feedback from the group to see if we have consensus on this
approach, i.e.:
1. draft-ietf-opsawg-discardmodel:
a. IM (YANG)
b. DM (YANG)
2. draft-evans-opsawg-ipfix-discard-class-ie
a. IM (references draft-ietf-opsawg-discardmodel)
b. DM (IPFIX IE)
The alternative would be to have a discrete draft for the information model,
e.g.:
1. draft-ietf-opsawg-discardmodel
a. IM (YANG)
2. draft-ietf-opsawg-…
a. IM (references draft-ietf-opsawg-discardmodel)
b. DM (YANG)
3. draft-evans-opsawg-ipfix-discard-class-ie
a. IM (references draft-ietf-opsawg-discardmodel)
b. DM (IPFIX IE)
Cheers
John
From: "Evans, John" <jevan...@amazon.co.uk <mailto:jevan...@amazon.co.uk>>
Date: Friday 22 November 2024 at 13:43
To: "mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>" <mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>>, Benoit Claise
<benoit.cla...@huawei.com <mailto:benoit.cla...@huawei.com>>, "opsawg@ietf.org <mailto:opsawg@ietf.org>" <opsawg@ietf.org <mailto:opsawg@ietf.org>>,
"opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>" <opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>>, Jeff Haas <jh...@juniper.net
<mailto:jh...@juniper.net>>, "Pylypenko, Oleksandr" <o...@amazon.com <mailto:o...@amazon.com>>
Cc: "draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>" <draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>>
Subject: Re: draft-ietf-opsawg-discardmodel
Resent from: <alias-boun...@ietf.org <mailto:alias-boun...@ietf.org>>
Resent to: <akad...@cisco.com <mailto:akad...@cisco.com>>, <jevan...@amazon.co.uk <mailto:jevan...@amazon.co.uk>>,
<jh...@juniper.net <mailto:jh...@juniper.net>>, <mohamed.boucad...@orange.com
<mailto:mohamed.boucad...@orange.com>>, <o...@amazon.com <mailto:o...@amazon.com>>
Resent date: Friday 22 November 2024 at 13:42
Hi Med / Benoit / Jeff / All,
I agree re option 2. If there are no objections we’ll refactor the draft on
that basis.
Cheers
John
From: "mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>"
<mohamed.boucad...@orange.com <mailto:mohamed.boucad...@orange.com>>
Date: Friday 22 November 2024 at 07:35
To: Benoit Claise <benoit.cla...@huawei.com <mailto:benoit.cla...@huawei.com>>, "Evans, John" <jevan...@amazon.co.uk
<mailto:jevan...@amazon.co.uk>>, "opsawg@ietf.org <mailto:opsawg@ietf.org>" <opsawg@ietf.org <mailto:opsawg@ietf.org>>,
"opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>" <opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>>
Cc: "draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>" <draft-ietf-opsawg-discardmo...@ietf.org
<mailto:draft-ietf-opsawg-discardmo...@ietf.org>>
Subject: RE: [EXTERNAL] draft-ietf-opsawg-discardmodel
Resent from: <alias-boun...@ietf.org <mailto:alias-boun...@ietf.org>>
Resent to: <akad...@cisco.com <mailto:akad...@cisco.com>>, <jevan...@amazon.co.uk <mailto:jevan...@amazon.co.uk>>,
<jh...@juniper.net <mailto:jh...@juniper.net>>, <mohamed.boucad...@orange.com
<mailto:mohamed.boucad...@orange.com>>, <o...@amazon.com <mailto:o...@amazon.com>>
Resent date: Friday 22 November 2024 at 07:35
Hi Benoît, all,
I think that I can convince myself that option 2 is better here (have both
IM/DM in the same spec). The main challenge will be to find “where” to anchor
the nodes (interface, NE, routing management, etc.). Otherwise, I expect the
structures/groupings we already have in the IM will be reused nicely. However,
we will need to register the IM as well as we need to import it.
I’m not sure to understand the last part of the following:
==
3. Should information models specified in YANG also be registered in IANA?
This is where we might have different views. As mentioned during the meeting: "If
you have it in IANA, then people will expect tooling to work.".
=
The tooling (pyang, etc.) still work even with the current IM in the draft.
Also, the module can be imported/augmented/etc.
Please note that even **fake** modules were registered! Think about
ietf-template for example:
==
ietf-templateietf-templ...@2010-05-18.yang
<mailto:ietf-templ...@2010-05-18.yang> N
urn:ietf:params:xml:ns:yang:ietf-template temp [RFC6087]
===
Thank you.
Cheers,
Med
De : Benoit Claise <benoit.cla...@huawei.com <mailto:benoit.cla...@huawei.com>>
Envoyé : jeudi 21 novembre 2024 18:22
À : Joe Clarke (jclarke) <jcla...@cisco.com <mailto:jcla...@cisco.com>>; Evans, John
<jevan...@amazon.co.uk <mailto:jevan...@amazon.co.uk>>; opsawg@ietf.org
<mailto:opsawg@ietf.org>;opsawg-cha...@ietf.org <mailto:opsawg-cha...@ietf.org>
Cc : Pylypenko, Oleksandr <o...@amazon.com <mailto:o...@amazon.com>>; Jeff Haas <jh...@juniper.net
<mailto:jh...@juniper.net>>; BOUCADAIR Mohamed INNOV/NET <mohamed.boucad...@orange.com
<mailto:mohamed.boucad...@orange.com>>; Aviran Kadosh (akadosh) <akad...@cisco.com
<mailto:akad...@cisco.com>>
Objet : Re: draft-ietf-opsawg-discardmodel
Dear all,
Joe and I spoke. As we see it, there are multiple questions here.
1. Is the IETF interested into standardizing information models?
As mentioned by Mahesh (in the OPSAWG meeting minutes, to be posted soon):
"generally we don't spend too much time on info models and work on standardise data
models". However, we believe, as we accepted the discard information as WG document
already, let's not revisit this decision.
2. How should those information models be specified?
Could be text, UML, whatever.
YANG is also a possibility (even if not common practice) as YANG is a data
modeling language. As John mentioned during the meeting, one argument in the
favor of YANG is that it's well known.
3. Should information models specified in YANG also be registered in IANA?
This is where we might have different views. As mentioned during the meeting: "If
you have it in IANA, then people will expect tooling to work.".
What do we do from here?
1. Information model published as informational document: YANG model in an
appendix, not normative. In such a case, you don't register the YANG module in
IANA.
2. Information and data models in a single standards-track document, with the
data model registered as a YANG module with IANA.
Number 2 might be better, as you mentioned that there are existing implementations.
Plus it might ease maintainability and give other implementations something to root
to and augment. This might also be the path of least resistance to publish. The YANG
model in IANA would also justify the fact that you moved from Informational to
Standard Track in this version
(https://datatracker.ietf.org/doc/draft-ietf-opsawg-discardmodel/03/
<https://datatracker.ietf.org/doc/draft-ietf-opsawg-discardmodel/03/>)
Jeff, at the microphone during the OPSAWG meeting, mentioned the issue of
evolving/maintaining information models before going to a data model. We are
not too sure how publishing an information model as RFC (as opposed to a WIKI
or something similar) is actually helping out.
Regards, Joe and Benoit
On 11/13/2024 11:19 PM, Joe Clarke (jclarke) wrote:
As a contributor, I think a data model approach would be far more useful. That
said, I haven’t seen these proprietary implementations based on your draft info
model to understand how they deviate or what data modeling approach they take.
I still think that starting with a YANG data model wouldn’t preclude future
drafts standardizing IPFIX-based approaches to packet discard reporting (just
as we’ve seen MIBs move to YANG)
