LISP WG

CHAIR(s):

• Joel Halpern ( jmh AT joelhalpern.com )
• Luigi Iannone ( ggx AT gigix.net )

SECRETARY:

• Padma Pillay-Esnault ( padma.ietf AT gmail.com )

Session 1/1 (60 Minutes)
Thursday, July 29, 2021
19:00 - 20:00 (UTC), Session I, 60 Minutes
[3:005(+1) - 4:00(+1) Beijing - Friday early morning!]
[21:00 - 22:00 Paris]
[15:00 - 16:00 New York]
[12:00 - 33:00 San Francisco]
Room 1: https://meetings.conf.meetecho.com/ietf111/?group=stir&short=&item=1

Agenda

1. Administration

• Halpern/Iannone

  • Agenda Bashing
  • Status reports for WG drafts
  • 5 Minutes (Cumulative Time: 5 Minutes)

• WG Status: in good shape only LISP-Sec and LISP-Nexagon are waiting for the shepherd's writeup.

  • LISP-Sec authors will update document early September
  • LISP-Nexagon: write-up to come in the coming weeks.
• Chairs will make explicit call on the mailing list to check interest on the remaining WG items and move them forward if there is still interest.

2. PIM Join/ Prune Attributes for LISP Environments using Underlay Multicast - https://datatracker.ietf.org/doc/draft-vgovindan-pim-jp-extensions-lisp/

• 10 Minutes (Cumulative Time: 15 Minutes) - Stig Venass
• Authors are looking for feedback from the LISP WG. The document will advance in the PIM WG. WG last call can be done in both WGs.
• CHairs: Seems OK to move forwrad this way.
• Q&A:
  • Are LISP nodes and non-LISP nodes allowed to join the PIM distribution tree?
  • It is allowed. There is not easy way to prevent it. Any router can join the tree, but the traffic is LISP encapsulated so non-LISP router may not know what to do with it. Ideally we should reserve some addresses and have an underlay dedicated to this purpose.
  • Future solution may include LISP signal free multicast.

3. - LISP Transport for Policy Distribution - https://datatracker.ietf.org/doc/draft-kowal-lisp-policy-distribution/

10 Minutes (Cumulative Time: 25 Minutes) - Michael Kowal

• Authors ask for feedback and comments (knowing that this dicument would depend on the LISP-name-ecoding draft on which there is already onging discussion.
• Q&A
  • Are you assuming this is a dedicated LISP system where the LISP mapping system is only being used for this purpose? Or these are actually distinguished names and if so what is the mechanism to make the policy names distinguished?
  • For this use case the aim is to be able to use a single distinguished name to identify a policy to be used on a group of xTRs on which an operator wishes to push such policy.
  • Names do not seems "distiguished" in the general case the authors are invited to comment on this point on the mailing list. Also, authors need to clarify differences with other policies mechanisms (e.g., PCE, BGP, SR) and how they work together (and if they do not work together it should be stated).
  • The proposed mechanism remains a LSIP mapping system using a specific IID (Instance ID) so it runs in parallel. The names in this IID need to be unique.
  • There might be MTU issues in the signaling because of the use of JSON. IP fragmentation should be used.

4. - LISP - Fix

10 Minutes (Cumulative Time: 35 Minutes) - Sharon Barkai
- Authors asking for feedback on this work (also presented in COINRG)
- Chairs ask to take it to the list because we are behind schedule.

5. - LISP Uberlay

10 Minutes (Cumulative Time: 45 Minutes) - Victor Moreno

• Authors will publish requirements in a draft and then explore applicability of DDT (or find alternative solution).
• Q&A:
  • When talking about multi-homing is intended to different ASes?
  • It is the case since planes will connect to several radio regions that may belong to different organizations and hence different ASes. More to be discussed on the mailing list.

6. - LISP Protocol Open Items: a Perspective from a Commercial Implementation

15 Minutes (Cumulative Time: 60 Minutes) - Marc Portoles Comeras

• Due do lack of time the presentation has been interrupted. The chairs acknowledge the content. There are some WG documents that need to move forward.