Minutes IETF114: lisp
minutes-114-lisp-00

Meeting Minutes Locator/ID Separation Protocol (lisp) WG
Date and time 2022-07-27 17:30
Title Minutes IETF114: lisp
State Active
Other versions markdown
Last updated 2022-08-04

minutes-114-lisp-00

LISP meeting IETF 114

  • Administration
    Halpern/Iannone

    • Agenda Bashing
    • Status reports for WG drafts
      10 Minutes (Cumulative Time: 10 Minutes)

  • Minute taker: Jordi Paillisse

Albert Cabellos: is there something left to do?
Joel H: almost done. RPC (RFC Production Center) will ask to review and approve + some questions. Please be responsive when they reach to you.
A. Retana: We requested expedite processing for the whole cluster of documents. Please reply fast to the RPC questions. They'll probably come together.
Dino: I'm worried about the intro document and Noel.
Joel: Noel is no longer a co-author.

[Luigi I. review pending docs]
o WG Items

Alvaro R: In case of lost state of reliable connection there is no specific place where to restart.
Marc P: The idea was to keep things simple in order to fall back.
Alvaro: But you are already authenticated.
Marc P: Agree, but you don't know when you've lost state
Dino: There is anyway periodic re-authentication.
Luigi I: Suggest to mandate at least one reliable transport, namely TCP. Order less important, no need ot mandate anything because choice is limited.
Dino: Need new UDP port for QUIC. For SCTP is the same like TCP and UDP?
Marc: The SCTP document has their way to request SCTP ports but we may request 4342 if available
Luigi: Just a matter of finding the right registry. Port 4342 should work for SCTP if it's not taken.
Marc P: For QUIC might be different.
Luigi: Agreed.

o Non WG Items

Dino: I want to request WGLC
Joel: You just want to go throught WG adoption and then we can issue WGLC without changing the name.
Dino: Already asked about WG adoption on the mailing list.
Alvaro: Please rename it before going to the IESG.
Joel: We will issue the adoption for call next week, 3 weeks, then change name, then WGLC, then to AD.
Luigi: Once adopted, we can ask for early review of the other areas.
Joel: Is a short document no need for early reviews.
Dino: What should be the intended track?
Joel: Can only be Proposed Standard.
Sharon B: I think it's a great idea. Solves problems using a routed namespaces. Select use cases to add to the charter.
Dino: (referring to slide 4) All these drafts are in the scope of the charter.
Marc P: You're missing LISP VPNs draft.
Dino: I didint' see the reference, we should add it.
Sharon: We should add use cases to the charter.
Luigi: The current charter focuses on the bis-documents. Once they are published be can start discussion for re-chartering and re-focusing the work of the WG.
Alvaro: We can re-charter between now and London.
Alvaro: About the document, you are not specifying how to use this.
Dino: The docs I have referenced explain the uses cases and how to use it.
Alvaro: This uses ASCII instead of UTF8 for instance. This question will be raised at some point.
Dino: I don't have an opionion either way.
Marc: Not necessary for now.
Alvaro: We're doing it for instance in BGP.
Joel: We can't go to UTF8, we'd need to redo the structure. We're using zero termination, does not work with UTF8.
Dino: Better leave it as is.
Alvaro: Just be ready for the question.
Joel: The question will come up and we'll have to answer.

Marc P: Have you thought about mobility?
Dino: Devices on the grund moving from ground station to ground station will use EID-mobility.
Marc P: Can the xTR move?
Dino: New problem. Satellites are moving and question is wheter they will assign new RLOCs to ground stations. This remains similar to EID mobility and all of the LISP machinery should work.
Tianji Jang: What is specific to satellite? This is the same like any other mobility scenario.
Dino: It's another type of underlay.
Tianji Jang: I cannot see the technical difference. Why is satellite relevant?
Dino: Its just another use case.
Erik Kline: Is there a presumption that the xTRs have terrestrial control plane connectivity all the time? Or they may have it over the orbital domain? If so, does that have implications on the control plane reliability?
Dino: We assume it is in the ground, but you could put the mapping system up if the satellite can route. But maybe less reliabile.
Erik: And also higher latency.
Luigi: (referring to slide 5) What about "across RLOCs in space"? Then the satellite is an xTR.
Dino: You load splitting across two ground station meaning accross RLOCs.
Mike K: Is there any MTU limitation between the xTR and the satellite? Military have encruyption in the front so have a hard time to make things like GRE go through.
Dino: I don't know yet.
Jordi P.: From a LISP perspective does it make sense to treat satellites like LISP mobile nodes?
Dino: So giving them an EID.
Jordi P.: Yes, since you have all these links continously changing.
Dino: Satellites will move around but they are not necessarily using IP addresses so they look like always being the same but just be momentarily unreachable. Usinf EIDs on the satellites is a different use case that I have not considered.
Luigi: Please continue on the mailing list. Tianji please send as well you comment to the mailing list.

  • Overflow Time/ Discussion
    15 Minutes (Cumulative Time: 60 Minutes)