Minutes interim-2021-lpwan-07: Tue 16:00
minutes-interim-2021-lpwan-07-202105041600-01
Meeting Minutes | IPv6 over Low Power Wide-Area Networks (lpwan) WG | |
---|---|---|
Date and time | 2021-05-04 14:00 | |
Title | Minutes interim-2021-lpwan-07: Tue 16:00 | |
State | Active | |
Other versions | markdown | |
Last updated | 2021-05-04 |
Meeting Information
Meeting links
- https://cisco.webex.com/cisco/j.php?MTID=m72aad28e8a2a7ce36076f9b09814f8d6
- CodiMD
- Session info
- Session recording
- Topic: LPWAN WG interim-20210504 1414-1
- Password: xEPAP72f
Laurent
Date / Time
- Date: May 4th, 2021
- Time: 7-8am US Pacific PST, 4pm CEST
Attendees / Blue Sheets
- Carsten Bormann, TZI
- Éric Vyncke, Cisco
- Pascal Thubert, Cisco
- Alexander Pelov, Acklio
- Laurent Toutain, IMT Atlantique
- Julien Catalano, Kerlink
- Carles Gomez, Universitat Politecnica de Catalunya
- Sergio Aguilar
- Vincent Audebert, EDF
- Olivier Gimenez, Semtech
List to copy from (people not here):
- Juan Carlos Zúñiga, Sigfox
- Diego Dujovne, UDP
- Georgios Papadopoulos, IMT Atlantique
- Dominique Barthel, Orange
- Ana Minaburo, Acklio
- Edgar Ramos, Ericsson
- Paul Duffy, Cisco
- Ivaylo Petrov, Acklio
- Arunprabhu Kandasamy, Acklio
- Sean Turner, sn3rd
- J Lecoeuvre
- David Millman
- Ricardo Andreasen
- Arthur Oval
Agenda
[16:05] Administrivia [15min]
- Note-Well, Scribes, Agenda Bashing
- WG Status, IETF 111 query
[16:20] SCHC Architecture [20min]
[16:40] Data Model for SCHC [15min]
[16:55] AOB [ QS ]
Minutes
Agenda
[16:05] Administrivia [15min]
- Note-Well, Scribes, Agenda Bashing
- WG Status, IETF 111 query
Alexander presents the intro, agenda bashing, AD bashing, IPR notes etc
Minutes of last meeting approved
Taking the pulse of the WG about having a WG meeting at IETF-111: usual suspects will not wake up after midnight for an LPWAN meeting
[16:20] SCHC Architecture [20min]
- new version
Carsten: please do record the meeting
Pascal: done (we are not recording all meetings by default)
rename to architecture/framework ? since it explains how to distribute keys. Explain the SCHC protocol can be deployed. What are the end-points.
ToC:
- discuss schc technology and implies that a profle is needed for technology
- what are the end points, different layers
- data model
Eric V: how the rule are sharing rules
Pascal: we will document that, data model says how it is install. But some text is missing now, work in progress. Can be improved.
Need more aggreement and discussion on the mailing list to converge,
2 types of end point : device and P2P
Security consideration; focuses on rules distribution
Eric: point attacks against encrypted traffic using dynamic dictionnary based compression (like ZIP) that are not effective to SCHC because of the static rules. This was raised during the IESG evaluation of CoAP/SCHC I-D.
Laurent will bring pointers
Pascal: Is there point that are missing ?
- do we add an applicability statement ?
Alex: it makes sense
Carsten: compress Goose?
https://en.wikipedia.org/wiki/Generic_Substation_Events
Pascal: lot of strings that can be compressed easily
Carsten: over LPWAN
Pascal: its generic it can be applied in other kind of networks.
Vincent: it is a L2 protocol, it is not a problem of size, more about routing
Vincent: Happy to work on it.
Pascal: can we plan an intermin meeting on goose ?
Vincent: yes, but not next meeting
Alex: there is a lot of protocol that can apply and 5G is also in our charter.
Pascal: applicability will grow, and may be a new document in the future
Pascal: Adoption ?
Eric: the group has to decide, this document is useful with the YANG DM.
Pascal: it gives the context, both are needed.
Alex: call for adoption. Nobody against, to be confirmed the mailing list.
Eric: will give the status since there is a conflict of interest.
[16:40] Data Model for SCHC [15min]
- Yang Doctors feedback
Lurent presents on the YANG doctors review. We're not experts many thanks to Carl Moberg.
New version on the repo, reformatted with PYANG. NOw WG doc.
review actions: different groups, use YANG feature to make parts mandatoy and others not.
Laurent wishes feedback. Do we need that?
Pascal: the protocol is very generic, support the idea to make everything feature