Minutes for SOFTWIRE at IETF-93
minutes-93-softwire-1

Softwire WG Agenda @ IETF 93
Monday, July 20, 2015
17:40-18:40, Monday Afternoon session III @ Congress Hall III
Chairs: Yong Cui & Suresh Krishnan

1. Agenda Bashing, WG & Document Status (Chairs) 10 minutes

Yong: This is the Notewell for IETF meeting and welcome our new AD, Terry
Manderson. Terry, would you like to say something?

Terry said hello to everyone.

Yong: For this agenda, we will have several discussions on YANG model, RADIUS
extensions and multicast support. After that we will discuss the future
direction of Softwire. For our WG documents, we have 5 WG documents in RFC
editor queue. -Lw4over6 and 4rd got approval from the authors -Map, map-dhcp,
and map-t still need authors' approval Currently we have 8 WG documents in
total. -3 MIB documents will advance to IESG next week We also have multicast
documents and others on MAP RADIUS and deployment, we will show our milestone
later.

2. YANG Data Model for IPv4-in-IPv6 Softwire, Linhui Sun / Ian Farrer - 10
minutes
   http://datatracker.ietf.org/doc/draft-sun-softwire-yang/

Ian: Split one module into 3 modules? (One per sw mechanism)

Suresh: We don't know if this is gonna stay in the same structure if it gets
adopted by the WG. Another way to look at it is combining the common part from
the three mechanisms. It's a bit pre-mature to decide on that.

Ian: Have you got specific suggestions on how that could be combined?

Suresh: No, the thing is like we've done some rationalization of the
provisioning solutions. Nothing stopping people from saying we can rationalize
this as well. Not certain if it's gonna be better but it's gonna be different.
I don't know yet because we haven't tested the consensus of the WG on that, so
put off the discussion on one or three until that point.

Ian: OK.

Ian: Adopt it as a WG item?

Suresh: So how many people in the room have read the draft?

Suresh: Not enough hands, will start something on the list.

3. RADIUS Extensions for IPv4-Embedded Multicast and Unicast IPv6 Prefixes, Wei
Meng - 10 minutes
http://datatracker.ietf.org/doc/draft-hu-softwire-multicast-radius-ext/

Wei: Adopt it as a WG item?

Suresh: Few people have read it. This thing requires a re-charter, we held a
discussion at the end. At this time, we are not able to adopt. But once we
finished the re-charter, if there're interests from operators we will certainly
consider it.

4. Multicast Support for Mapping of Address and Port Protocol and Light Weight
4over6, Behcet Sarikaya - 10 minutes
http://datatracker.ietf.org/doc/draft-sarikaya-softwire-map-multicast/

Behcet: Now discuss what can be done, I believe it's important. We need to
progress the draft to a better version, I will leave it to the chairs.

Suresh: The same answer like the way before, need to see if it meets deployment
demand.

Behcet: Map is one of the most important protocols developed. Need multicast in
map network. Lw4o6 is improved version of DS-Lite, I think this multicast
approach is gonna be needed.

5. Re-charter (Co-Chairs)

Graceful shut down around Buenos Aires meeting (April 2016)
Important items: Unified CPE and deployment drafts

Suresh: Continue work on the Unified CPE draft?

Ida (jabber): Yes, we should finish the Unified CPE document.

Ian: Are the operational documents referred here the same as the ones discussed
in v6ops

Suresh: They are disjoint at this point. The question is do you want these
documents? Could be done in v6ops, but will still get last call in Softwire.

Ian: What I'm planning is to see what exists at the moment. See what we have
from our own operational experience. See if there is anything meaningful we got
to say about.

Regarding the Unified CPE, we've done a lot of work on testing the OpenWRT
version of this in our environment. We are running DS-Lite and
Lightweight4over6. We need to swap between those modes in the lifecycle, but it
doesn't work. This is undefined behavior. We are getting to the stage where the
loose ends are starting to matter. Unified CPE is one of those. Another draft
is the multiple tunnel endpoints draft, about differences between how the
implementers have chosen to build the AFTR. Looking back at the lightweight
spec, it's not something we covered.

Suresh: We will submit a new charter after we talk to the AD, along with the
proposed milestones to the WG for discussion. If you think there's something
missing, please speak up.

Ian: I think we need tidying up in a few areas to make sure we got complete
working solutions.

Suresh: It would be good if you can spend effort on the Unified CPE draft to
update it.

Wei: I didn't see the radius for multicast in the proposed milestone. What does
the WG think about the multicast radius?

Suresh: When we propose the new charter, you can say you want it to be added
and we can discuss on the list. We proceed to see if we add it or not based on
consensus.

Terry: There is a large number of items which need to get through. I'm worried
about the amount of people in the WG who are going to contribute to that.

Suresh: There's going to be 4 new documents. Other documents have been reviewed
or last called already, the working group's job is done, we (as chairs) are
holding the documents.

The Yang model, the deployment models, and the Unified CPE need significant
reviews, we are questioning whether we need them or not.

Terry: Make call on the mailing list to ask whether there are people willing to
spend time on those documents. That will give indication on whether to continue
or call those documents dead and close working group.

Chongfeng: I hope Radius Extension for IPv4 Multicast can be added to the
milestone. It's been discussed many times, and it's a real requirement from
China Telecom.

Suresh: Feel free to comment when we send out the charter for review. We need
people reviewing the document, probably someone with multicast knowledge.