# ROLL session at IETF 112 meeting

Meeting Material: 

https://datatracker.ietf.org/meeting/112/session/roll

## Agenda:


* 14:30 - 14:45  (15 mn) Intro, WG Status  (Ines/Dominique)

* 14:45 - 15:00  (15 mn) AODV-RPL MOP (Ines/Dominique)

* 15:00 - 15:15  (15 mn) draft-thubert-6lo-multicast-registration (Pascal)

* 15:15 - 15:29  (14 min) Open Issues
  -  Reviews on DAO-Projection: WGLC? (5 min) Pascal
  -  draft-iwanicki-roll-rnfd adoption? (5 min) Konrad
  -  Reviews on enrollment-priority: WGLC? (4 min) chairs

-  15:29 - 15:30  (1 mn) AOB (Everyone)


# Action points

* reviews needed for Root-Ack
* milestone dates to be updated (chairs)
* DAO projection: start WGLC after reviews from Michael, Rahul and Aris are received and processed.
* chairs: issue a warning on the ML that we are intending to reuse MOP 4 for AODV, soliciting reactions.
* MCR: evaluated consequences on existing RFC's of reuse of MOP 4 for AODV.
* Dominique/Ines to talk to 6lo chairs about ROLL also reviewing/LCalling 6lo-multicast (because of ROLL content)
* Pascal to inform PIN of 6lo-multicast-registration
* chairs to ask the mailing list about adopting RNFD, discuss experimental / standards track.
* Enrollment-priority: authors to confirm among them that min-priority is unchanged down the DODAG, post new rev, chairs to get one more review than WGLC.

# Meeting notes (times in UTC)

Note takers: Michael Richardson, Pascal Thubert, Rahul Jadhav

[14:30]

## Intro, WG Status

Note-well, notetakers.
Ines browses the I-Drafts at the WG.
A new version of the root ack was published, not discussed today, please review !.

+ Milestones
Need to update the dates.
  * For MOPEX: Rahul has one open question, only open action item. With that done, ready.
  * Capabilities will take longer.
  * (in the chat) MCR suggests IETF114 for Cap and Mopex
  * DAO projection needs more reviews (Michael and Remous Aris).
  * DIS-modification: submission end of 2022
  * Currently no focus on BIER and CCAST so this is put on hold.

+ Tickets open
Both in Github and in the Tracker, moving to Gihub.
Authors: please go look at your tickets.

[14:41]

## AODV-RPL 

Version 11, DISCUSS comments from IESG being resolved.

Using MOP4 would free MOP 5, only codepoints 5 and 6 left until we have MOP extension.

Obsoleting RFC 6997 would free MOP4, or we could have both protocols use MOP4.

Pascal: overlap will not happen by accident, need to configure L2 keys, OF, etc...

Pascal: obsoleting 6997 is ok. Having 2 solutions on the market at the same time blurrs the picture. 6997 was experimental, this one is going standards track, that's exactly what we mean.

Alvaro: applicability RFCs recommends using P2P (SHOULD or MUST?), will need an update.

Alvaro: do parts of AODV use parts of P2P? in which case, can't refer to an obsoleted document, need to specify them in extenso in the document. We need to be careful.

MCR: "Building and Home Applicability" RFC7733. Not implemented in significant amount, afaik. Hopefully saying this updates 7733 should suffice, a paragraph to indicate how this replaces 6997.

Alvaro: maybe just a sentence.

MCR: will check the applicability and provide review if there is any overlap between 6997 and AODV-RPL.

Alvaro: AODV-RPL already went through IESG review. Having complex issues that might need recirculating in WG. We'll discover the scope of the changes before we decide whether to pull the document back from IESG.

Dominique (chair hat): discussion on the ML leaning towards reuse of MOP 4. Any strong objection here? 

No objection heard.

Dominique: we'll issue a warning on the ML that we are intending to reuse MOP 4, soliciting reactions. In parallel, will do the homework to evaluate what extra work this entails.


[14:52]

## draft-thubert-6lo-multicast-registration

Proposed at 6lo but this is both a 6lo/ROLL work.

There is support for RPL multicast with storing mode, but nothing for non-storing mode.

Needs a new MOP for this. If AODV uses MOP5, only MOP6 is left.

Alvaro: this eliminates need to do MLD?

Pascal: pretty much. MLD is much complex.

Eduard V (on chat): This type of multicast on the last hop has a big pre-request: replace normal ND. Unfortunately, not many people would agree.

Alvaro: 8505 and related could be used in any network, not just RPL.

Alvaro: MLD moving into an internet standard. Please inform PIM, send an email, to make them aware of this.

Alvaro will talk to Erik Kline.

Pascal: BESS would also be interested.

Alvaro: all we're doing here is multicast targets, not multicast sources, correct?

Pascal: yes, and this this is source-independant only.

MCR: should this doc be split into a 6lo doc and a ROLL doc? 6lo might show no interest because they think it's dependent on ROLL.

Alvaro: important for ROLL to look at this document as well. Fine for 6lo to host this document.

Alvaro: Dominique/Ines to talk to 6lo chairs to make sure we Last Call this document here as well.

MCR asks how to get the non-ROLL parts out into the wider community? INT-AREA? How can we start a move towards not requiring MLD to make multicast work on switched ethernet systems?

MCR: Move away from Layer 2 tricks to use layer-3 only.


[15:15]

## Reviews on DAO-Projection

Pascal believes that doc is ready for WGLC anytime.

Understand Aris and MCR still committed to do reviews.
  - (chat) MCR will do review by end of next week. WGLC in Dec?
  - (chat) Rahul will provide a review next week as well

Close tickets? Pascal says they are very old and obsolete.
  - Pascal to setup a dedicated work meeting with the chair(s) to resolve this


[15:21]

## RNFD  - Fast border router crash detection in RPL


Applied comments from WG. Major comments addressed.

Pascal: You have to find an agreement at least between 2/3rd of the first hop 6LRs, and you are want them to communicate without going through the root, which makes the problem even harder. 

Pascal: agreed that the node needs to know if the root is still there. And they need to do it rapidly. There are few things in the document that we are not completely sure. Yes to approve to adopt .. but not a final say on the algorithm.

Konrad: Would be fine. Could correct some of hte algorithms, might make sense to get rid of few parameters, etc. 

Dominique: Adopting means the WG agrees to work on the topic. Not an acceptance of solution.

MCR: adopt a problem statement not the solution.

Dominique: Any objection here to adopting this document?
No objection heard. We'll confirm on the mailing list.

Pascal: if we adopt .. we need to decide whether it will be experimental or standards track.

Dominique: Not sure if we need to decide about that now. Only needed before IESG submission.

Pascal: It would be good to know anyway.

MCR (on chat): too soon to decide whether standards vs experimental.


[15:27]

## Enrollment-priority 

MCR: authors should probably get together to discuss .. 

MCR: step back and try again with capabilities or something, I don't know

Pascal: I think is is perfectly shippable. WiSun is asking for it. Min priority that we are passing should be changed or not on the way down is something that needs to be fixed/clarified. Going down unchanged is perfectly workable.

MCR: I dont think we got that clearly in the doc in the last rev. Not sure authors are in consensus about no min-priority change.

Pascal: I think we were in consensus. As long as the min priority is unchanged in DIOs down the DODAG we have consensus and it is very workable.

MCR: Then may be we need to just ship. Will get a new rev posted by the end of this month.

Ines: We need one review more then go in WGLC. We need the reviews.


[15:32]

## Meeting is adjourned