Skip to main content

Minutes IETF117: madinas: Mon 22:30
minutes-117-madinas-202307242230-00

Meeting Minutes MAC Address Device Identification for Network and Application Services (madinas) WG
Date and time 2023-07-24 22:30
Title Minutes IETF117: madinas: Mon 22:30
State Active
Other versions markdown
Last updated 2023-07-26

minutes-117-madinas-202307242230-00

IETF 117 MADINAS WG Agenda

MAC Address Device Identification for Network and Application Services

Monday, July 24, 2023

15:30 - 17:00 Monday Session III Golden Gate 7-8

  • Welcome, Agenda Review and Status Update (WG Chairs)

      Reviewed the agenda, no comments
    
  • MAC Address Randomization current state-of-affairs (Presenter Carlos
    J. Bernardos)

      https://datatracker.ietf.org/doc/draft-ietf-madinas-mac-address-randomization/
      Daniel Borges: Asked if they tested macOS.
      Carlos: Clarfied that they focused on mobile OS so no.
      Juan-Carlos Zuniga: mentions that these documents are just a snapshot of the implementation.
      Warren Kumari: This document is stable, should we have a call for adoption.
    
      Took a poll for last working group call which 21 raised hands, an 3 non-raised hands.
    
  • Use cases and Problem statement (Presenter Jerome Henry)

      https://datatracker.ietf.org/doc/draft-ietf-madinas-use-cases/
      Juan-Carlos Zuniga: The working group hasn't considered which of these solutions 
      might be helpful.  Need to discuss DHCPv4/v6 and SLAAC best practices.
      Erik Vyncke: Draft seems to be heavy on wireless and lighter on DHCP, BGP
      and other IETF protocols.
      Jerome Henry: Agrees draft can improve with more then wireless information.
    
  • Status of IEEE 802.11 (Presenter Mark Hamilton)

      Juan-Carlos Zuniga: Suggest discussion happen in the open discussion part of the agenda.
    
  • WBA OpenRoaming Wireless Federation (Presenter Bruno Tomas)

      https://datatracker.ietf.org/doc/draft-tomas-openroaming/
      Daniel Borges: Thinks using OpenRoaming at the IETF would be a successful experiment.
      Erik Vyncke: IETF discussed using OpenRoaming for IETf-117, but the IESG decided not too.
      Juan-Carlos Zuniga: Wasn't clear if the IESG understood the purpose of the experiment.
    
  • Experimental results from OpenRoaming tests (Presenter Warren
    Kumari)

      Warren Kumari: CUI was the same on both the iPhone and iPad.
      Alan DeKok: CUI should be different per session, it's a bad implementation decision.
      Alan DeKok: This probably a failure of the documentation.
      Clemens Schrimpe: Do you have any results for Apple Watch?
      Warren Kumari: No, i can look into this a later date.
      Daniel Borges: Were the experiments with iOS 16?
      Warren Kumari: Yes
      Daniel Borges: Where do we fix the CUI problem? Radius/OpenRoaming?
      Warren Kumari: We used the radius proxy, and it's send the Identifier.
      Mark Grayson: OpenRoaming says CUI should be when privacy set.
      Alan DeKok: RFC 4372, says CUI should be temporary long enough to be useful. This should be clarified in the OpenRoaming document.
      Warren Kumari: It may change over time, weeks or months.
      Juan-Carlos Zuniga: Happy with the issues uncovered with this experiment, it's what we wanted trying to run it on IETF network but at a bigger scale, with more implementations, more IDPs, etc.
      Joe Clarke: Likes Alan's comment about fixing this. Looking into how to add different identfications. 
      Warren Kumari: Noticed the user ids might be tied to SIM numbers.
      Mark Grayson: A Hash will be used so this shouldn't be a problem.
      Clemens Schrimpe: How do we think about verifying people are using the standard.
      Alam DeKok: Specifications need to written to clarify what the intention is an how to implement it.
      Daniel Borges: Where other IDPs used in this experiment?
      Warren Kumari: We could create test content for verifying different IDPs.
      Daniel Borges: Who is doing the testing?  OpenRoaming?
      Bruno Tomas: WBA has legal contracts in place to resolve issues.  PKI allows for removing bad actors in case there are bad practices or abuses.
      Mark Grayson: According to the legal terms in WBA agreements, PII can be used for troubleshooting, but it can't be explioted.
      Erik Vyncke: Thanks to everyone for the interesting discussion.
    
  • Next Steps (WG Chairs w/AD Support)

      Juan-Carlos Zuniga: Looking for a Iterim Meeting to continue this discussion in about a month.  Watch the list for more information.