RAW WG Agenda IETF-116 - Yokohama

Date: Friday, March 31, 2023 - Session II
Time: 12:00-13:30 Tokyo / 4:00-5:30 UTC -- 90 mins

Chairs:
Rick Taylor
rick@tropicalstormsoftware.com

Eve Schooler eve.schooler@gmail.com

Responsible AD: John Scudder jgs@juniper.net

Onsite tool: https://meetings.conf.meetecho.com/onsite116/?group=raw
Meetecho: https://meetings.conf.meetecho.com/ietf116/?group=raw
Session materials: https://datatracker.ietf.org/meeting/116/session/raw

Shared notetaking: https://notes.ietf.org/notes-ietf-116-raw
Zuilip https://zulip.ietf.org/#narrow/stream/raw

Available post session:
Recording: http://www.meetecho.com/ietf116/recordings#RAW

Notetakers: Eve Schooler

1) Intro - Chairs -- 10 mins

• Administrivia
• Document status

Eve Schooler: LDACS published as RFC 9372!!! Major Congratulations to
all involved.

2) Use Cases - Carlos Bernardos -- 20 mins

https://datatracker.ietf.org/doc/draft-ietf-raw-use-cases/

John Scudder: next step will be for me as AD to review the updates
(particularly since Roman cleared all the Discusses). It looks like
there shouldn't be any further process needed. It that is the case, then
I approve, and off it goes to the RFC editor. Should be able to take
care of that in the coming week.
Carlos Bernardos: Thank you for all your support!

3) RAW Mobility - Carlos Bernardos -- 10 mins

https://datatracker.ietf.org/doc/draft-bernardos-raw-mobility/

Carlos: Mobility likely relevant to a number of use cases. Where the
terminal is mobile, but with constraints/requirements due to RAW.
Also presented this material in the DMM WG (distributed mobility
management).
Extensions to IPv6 for handovers. Not completely new on its own.
Collecting feedback from DMM and RAW.
Question: Do we think it is an interesting problem?

Janos Farkas: Is this technique needed in 5G, as a RAW technology?
Carlos: Not sure with the current specification if this has been
considered, with the very high constraints of RAW. Not sure how done
with URLLC.
Also in WiFi, depending on the technology, the optional step mentioned
in the Figure (step 0) may have more or less info. So it may be
something to integrate into the next version of the doc.

Rick Taylor: Whether this has applicability across multiple lower
layers? If yes, then it will be useful.
Carlos: We may also have, not explicitly mentioned yet, multiple
technology handover. A terminal may be attached to a WiFi pt of
attachment, gets out of coverage, then need to migrate to 5G. May need
something on top of L2 to support.

Balazs Varga: Re 5G, in 3GPP R18, they have generalized TCS support,
where the endpoints are fully covered with the functionality including
handover.

Carlos: Feedback so far is that there is interest.
Rick: yes, there is interest. Please continue.

4) RAW OAM -- Carlos Bernardos -- 10 mins

https://datatracker.ietf.org/doc/draft-ietf-raw-oam-support/

Presenting on behalf of co-authors.
The document is pretty stable, ready for last call.
About to be done both with DetNet and RAW OAM.
Eve: Likely need the architecture document to be stable before finishing
the OAM draft.
Rick: The doc has sufficient review. WGLC is a valid request.
Lou Berger: Going to last call is a good idea. Could ask: should this
wait for the Architecture doc, or what part is blocked by the
architecture? The OAM suite of docs: the DetNet framework is in John's
queue and we have an MPLS doc and an IP doc is almost there (one section
needs revision). These are all DetNet OAM documents.
John: Natural time
Lou: Good time to respond to the Architecture readiness. We need to
close off on those discussion issues from last October and make sure
they are addressed. Some are terminology. Some are how to represent the
wireless layer to the routing layer.
Rick: we will return to the Architecture WGLC topic later in the
session.

5) Discussion/Open Mic -- 30 mins

IEEE 802 Plenary in July!

Invitation from IEEE 802 to provide a RAW Tutorial at the upcoming
plenary in July.
Janos will provide the DetNet background, and Carlos to provide the RAW
overview.
Janos in chat window: Remote participation at the IEEE 802 tutorial will
be possible for free. Information will be available here:
https://www.ieee802.org/Tutorials.shtml

Changes to RAW WG:

Eve: Eve and Rick have had a change of work circumstances and are
needing to step-away from co-chairing RAW. After discussion with John,
Lou, Janos, the idea is to re-integrate RAW back into DetNet, from which
it was spawned.
Rick: The intention however is to shepherd RAW through the transition.
Have been a bit stuck with the architecture doc, and think it would
benefit by pulling it into the broader DetNet forum, to achieve wider
review.
It is a super valuable piece of work that could use scrutiny beyond RAW
which has turned out to be a rather smaller venue.

John: If folks have concerns, please reach out to the chairs (RAW or
DetNet) or to John or to whomever you are most comfortable, and we will
happily consider your inputs.

Carlos: Thank you to the co-chairs.
Question: how do we plan to add into DetNet, which already has a full
plate?
Rick: Have been talking with Janos and Lou about exactly that.
Lou: Echo thanks for shepherding the activities in RAW. How are we
thinking of managing the added load? DetNet is in a different place than
when RAW was spawned. A lot of the original work is coming to a close
right now. The DetNet group is more mature, though it is not idle. We
have a new theme on queueing, which appeals to just a small segment of
interested parties in DetNet. The other parts of the group are more
about general deterministic networks, that can be leveraged to align
with RAW. We often manage the time by having side meetings that are
lengthy and that go deep on a topic. Overall: DetNet co-chairs are very
hopeful about the plan.

Rick from chat: Agree with John - this is not intended to be a sudden,
undiscussed transition. We want to make sure any transition is good for
RAW and happens on a sensible timescale.

Poll: If you only attend RAW, raise your hand. About half of the RAW
participants on the meeting ONLY attend RAW vs RAW+DetNet.

Rick: Key concern: the future of the Architecture draft. It is a bit
stalled.

Discussion in chat window on the Architecture doc:

Pascal: I posted 2 revisions since last IETF, addressing terminology and
other issues.
Lou: I don't think my comments from October ever got addressed.
Pascal: I believe whatever is left can be WGLC issues.
Lou: I should say fully addressed, i.e., we still had some open points.

Pascal: Well we need a new review cycle to see what’s left. WGLC would
be fine for that
Pascal: I think I addressed the Overhearing thingy in January.
Janos: Now that IETF 115 was mentioned, I'm not sure that the IETF 115
discussion items have been addressed. For instance, PAREO as such is
layer violation, hence the draft does not follow the layering model.
Pascal: There were changes in the doc for t hat. Did you check the
diffs?
Janos: Well, PAREO is still in the draft...
Pascal: Yes and the interactions of layers discussed.
Pascal: the big PAREO section was removed as agreed.*

John: Agree with Pascal that sometimes the only way to drive the level
of engagement necessary to get a doc over the line is to declare a WGLC.
Let's suppose we'll use a WGLC as an announced period of time, "speak
now or forever hold your peace". Even if moving into DetNet, Eve and
Rick should likely be involved in that WGLC.

Lou: From an administrative perspective, do not have an objection to a
last call in BOTH working groups. From the technical side: the
disconnect in the discussion is because of the need to have a wider
perspective on the Architecture. For example, how the RAW architecture
fits into the existing DetNet and TEAS architectures. But the way it is
written is as if some of the concerns are narrowly just wireless, but
actually not unique to wireless. How do we close on those points? Both
terminology and technical, e.g., promiscuous over-hearing not just
wireless (Ethernet).

Rick: This was the outcome of the previous IETF. Agreement that wider
review still needed, beyond RAW perhaps.

Eve: Yes to the idea of WGLC in BOTH groups.

Lou: Would either of the RAW co-chairs be willing to be Shepherd?

Action items:

• Use Cases to final review by John, in short order, possibly as soon
  as next week
• WGLC OAM draft
• WGLC Architecture draft - jointly with DetNet - Eve to serve as
  shepherd
• Technologies - Carlos writeup forthcoming on the order of weeks
• Industrial Requirements - new contributors solicited to help it get
  to WGLC (Carlos, Corinna, Eve all interested)
• Ask the WG mailing list, who ONLY attends RAW vs RAW+DetNet