Minutes IETF125: dmm: Mon 01:00
minutes-125-dmm-202603160100-00

DMM Working Group - IETF 125 Meeting Agenda
Monday, 16 March 2026, 09:00-11:00

1. Administrivia

Luis as note-taker.

1. WG Status and Action Items (from IETF 124 Montreal)

John K: quick question on UDP tunnel-acaas. Do we need to present
something in the session?
Saturu: good to have a peesentation to understand if adoption is
succesful.

1. Working Group Drafts

• Mobile Traffic Steering
  draft-ietf-dmm-mts
  Marco Liebsch

Sri: TopoLoc what actually implies vs GeoLoc?
Marco: nodes that move could change in terms of both topological and
geographical location
Sri: my question is what is teh data plane
Marco: what is teh data type is kept yet open. If there is any service
that can provide more information (e.g., for satellites) we just provide
here a pointer without entering into details. Use cases is nto only for
future generation but also those that provide dynamicity.

John: questions for clarification. I'm also participating in 3GPP data
plane discussion. In a session there could be multiple flows. If we want
to reselect this use rplanes, which flows should be selected, what are
teh principles if we need to send different flows to different DCs?
Marco: one flow traverses one UPA: If the UPA changes is not a decision
of teh transport network but by the MCS. It is necessary that the MCS
notifies to the transport network. Tranprot network needs to steer the
traffic to the new UPA, and this is notifies by this API.
John: it would be good to have such clarification in teh document.
Second quesiton. How this impacts on flows above? Change sin IP
addresses? latency, queueing? All of those things impact and would be
good to have a note on what aere teh implciations of this. There are
some costs on what you are assuming.
Marco: ok

• A YANG Data Model for Attachment Circuit as a Service with UDP
  Tunnel Support
  draft-jlu-dmm-udp-tunnel-acaas-02
  John Kaippallimalil

Lionel: The content of the YANG model is what we did more than one year
ago. Good to keep adopted and with that move it forward to last call,
since it is the same content we have in the past.

Erik: reminder that chairs can move it forward after adoption

1. Individual Drafts
   Mobility Management for Inter-SNO Sharing in LEO Satellite Networks

   draft-lai-dmm-sno-collaboration
   Yunan Hou

John: The point of presence to internet is the same for both of the
independen satellite networks. Taht is a big assumption. Probably they
would be different. Taht could have implications on the architecture.
Another comment is: in the MNO case there is no ??, that is another big
assumtion. You need to think about that. In the MNO there is no
handopver betwene operators during the life of the session.
Yunan: LEO networks have fundamental differents with respecto to MNO. In
LEO network teh roaming is enforced.

Lionel: some working assumptions have to be validated. Specially the
need for inter-SNO. You will swith from one satellite to anotehr what
you will not change the IP address. The working assumptions need to be
evaluated and check for the need of IP handover on satellite network.

Saturu: how much of the cost would be shared? Do you have such costs
estimates?
Yunan: more than 9000 satellites.
Saturu: with the SNO 40% of the satellite will be reused, correct?
Yunan: 40% of the satellites is enough for full Earth coverage. By now
we only focus on performance, the cost will be part of future analysis.

Sri: focus first on the problem statement instead of jumping directly in
the solution.

1. AOB (Any Other Business)

Erik stepping down as AD, thanks so much. Tommy taking over the role of
responsible AD.