Clone wiki

meetings / 141024_webex

Minutes Webex 24 October 2014, 6TiSCH WG

Note: timestamps in PDT.

Connection details

Resources

Taking notes (using Etherpad)

  1. Pascal Thubert
  2. Xavi Vilajosana
  3. Thomas Watteyne

Present (template)

  1. Ariton Xhafa
  2. Diego Dujovne
  3. Guillaume Gaillard
  4. Giuseppe Piro
  5. Ines Robles
  6. Maria Rita Palattella
  7. Michael Richardson
  8. Nicola Accettura
  9. Pascal Thubert
  10. Pat Kinney
  11. Raghuram Sudhaakar
  12. Thomas Watteyne
  13. Xavi Vilajosana

Action Items

  • Thomas to update draft-ietf-6tisch-tsch by cut-off date.
  • Pascal and Thomas to talk to Ted about whether to push for 6TiSCH-specific CoAP draft, or wait for generic constrained management solution.
  • Xavi to cite draft-thubert-6lo-rpl-nhc-02 in next revision of draft-ietf-6tisch-minimal
  • Michael to publish a revision of the security draft with a clear text describing the needs in terms of data representation elements so the 6top data model can be updated.
  • Michael to publish a revision of the security draft with a clear text describing the needs in terms of data representation elements so the 6top data model can be updated.

Agenda

  • Administrivia [2min]
    • Approval agenda
    • Approval minutes last call
  • IETF91 [5min]
    • meetecho remote presentation
    • draft agenda
  • draft-ietf-6tisch-tsch-02 publication [2min]
    • please review
  • wrapping up draft-ietf-6tisch-minimal [15min]
  • presentation draft-richardson-6tisch--security-6top-02 [15min]
    • RPL option status and update 6lo [10min]
  • RFC7322: RFC Style Guide [5min]
  • AOB [5min]

Minutes

  • [08.05] Meeting starts [Chairs]
    • Pascal starts recording
    • Thomas presents agenda
      • update on IETF91 and draft agenda
      • present 2 draft and its status/evolution
      • Michael will present security draft status
      • Pascal will present status of RPL option at 6lo
      • if time permits, RF7322 will be introduced.
  • [08.08] Administrivia [Chairs]
    • Approval agenda

      No issues raised. Agenda approved.

    • Approval minutes last call
    • No issues raised. Minutes last call approved.
    • Thomas There was a security call on Tuesday. Went through draft-richardson-6tisch--security-6top-03. It will be presented today.
  • [08.12] IETF91 [Thomas]
    • IETF91 agenda is final. Meeting is 9am in Hawaii, i.e. 8pm CET time same day.
    • Meetecho available. Remote presentations and remote attendance.
    • Recording and minutes will reflect remote participation.
    • Michael Visa's from 104 attendees from China have not been approved yet.
    • Thomas First time that every session is in meetecho.
    • For remote presenters, there will be tests to make sure everything works well.
    • This Monday is the cut-off date for draft submission. Authors are called to submit their drafts before Monday.
    • Monday is also cutoff for draft agenda.
    • Initial agenda for 6TiSCH WG meeting.
      • present latest changes on drafts
      • TSCH draft, 6top interface draft, and minimal
      • introduction to plugtest at 93rd IETF meeting, with help of ETSI. Renamed from plugfest to plugtest as it related to verification.
      • Rechartering discussion to define next steps (include dynamic scheduling, etc.)
      • Security: Michael will present a revision of his draft and then RenĂ© has remarks

        Action: Thomas to update draft-ietf-6tisch-tsch by cut-off date.

      • Maria Rita volunteers to present TSCH draft
      • Xavi volunteers to present minimal and 6top interface drafts, if necessary
      • Both presenters will do it remotely.
  • [08.19] draft-ietf-6tisch-tsch-02 publication [Thomas]
    • version 02 published
    • follows the rewording proposed during the call
    • 3 issues still open:
      • http://tools.ietf.org/wg/6tisch/trac/ticket/25. Rewording proposal. Rene did not agree. There is a follow up at the ML.
      • an issue is opened to change mote to node as node is what we agreed.
      • Pascal what about term "LLN nodes"?
      • Thomas suggests to use term "LLN node" in first occurrence and follow with "node".
    • Pascal several draft are going to be sent to the IESG in November. The last call will be announced at IETF91 and then we will have an evaluation/call for consensus for 2 weeks before submitting.
    • List of drafts considered: TSCH, Minimal, 6top-interface, CoAP. We need to decide what we do with them.
    • Thomas what about architecture draft? What is the current status and what is the timeline for this draft.
    • Thomas Would there be a problem if we recharter without having finalized the architecture draft?
    • Pascal I do not see any specific problem.
    • Thomas what about the CoAP draft? what should we do? Wait for the COMAN activity with constrained RESTCOnf draft? or push for our CoAP draft.
    • Pascal we have milestones with the IESG. We can delay the CoAP draft after recharter to see what is the progress of the RESTConf, etc.
    • Raghuram Agrees to wait.

      Action: Pascal and Thomas to talk to Ted about whether to push for 6TiSCH-specific CoAP draft, or wait for generic constrained management solution.

    • Michael asks if there is a date for the ETSI event.
    • Thomas there is no specific date but it might be a couple of days before (or Monday) talking to IETF to see if this is possible. Target is July 2015 in Prague right before IETF93 is the target.
    • Michael the Contiki people planned to have a conference, hack fest right before. This might be a conflict.
    • Thomas the ETSI test event is not a long thing it is just a set of schedules of tests and people attend to their specific tests.
    • Ariton Regarding the interop. Are there going to be a set of scenarios?
    • Thomas yes. The plan is to develop the test scenarios and come out with a set of tests (end March)
    • Ariton there will not be certification on that event.
    • Thomas not sure. We need to check with ETSI.
    • The point is to make standard better. the outcome is to validate to our documents are correct and we are "optimal" in their design.
  • [08.xx] wrapping up draft-ietf-6tisch-minimal [Xavi]
    • change in the CCA option. Reflected in the drawing at the beginning of TX slot, timing before CCA. This is a clarification to follow exactly the standard
    • added the missing timing
    • open questions, need opinion
      • HbH header compression. Agreed last time to go to 6lo. Should not limit further extensions even at cost of additional byte?
      • what should minimal specify for security?
      • asking implementors if IE we are sending are enough?
    • Pascal: published draft-thubert-6lo-rpl-nhc-02
    • offers 3 possibilities:
      • (greedy) the one from NHC: consume 1/4 of NHC space (64 possibilities used)
      • (conservative) don't change layout at all, enumeration in 6LoWPAN NHC already, consume new header (but then need full byte after it) (1 possibilities used, but extra byte)
      • (middle) idea from Carsten: limit the same to something the same as greedy, but add a byte when "error". Consumes 4 bits (~20 possibilities used, 1 extra byte when error)
    • Thomas what is an "error"?
    • Pascal forwarding error (O,F bits)
    • Pascal problem is that you might need to change packet length during forwarding
    • Thomas agreed that this is the draft to cite in minimal?

      Action: Xavi to cite draft-thubert-6lo-rpl-nhc-02 in next revision of draft-ietf-6tisch-minimal.

    • [Xavi] what should draft-ietf-6tisch-minimal contain in terms of security?
    • Pascal I do not see the need to say anything, RPL has its own security, any security that applies to other protocols apply to minimal.
    • Michael similarly, there must be L2 security, which we just inherit.
  • [08.57] presentation draft-richardson-6tisch--security-6top-02 [Michael]
    • Missing text for security pieces in 6top
      • join controller provides a certificate for the node (500 bytes), node may issue a cert request.
      • OR (red line at the bottom) 802.15.9 key exchange, per peer keying using group key. Large shared secret like 32 bytes. Are they mutually exclusive?
    • security pieces for the 6top model. What do we need to configure specifically?
    • the join controller provides the certificate.
    • the security for the DTLS connection needs to be able to identify the used certificate and its location.
    • the certificate will be used by p2p exchange. Also there can be a group key shared by all nodes.

      Action: Michael to publish a revision of the security draft with a clear text describing the needs in terms of data representation elements so the 6top data model can be updated.

    • With a PCE each node needs to have a secure relation with the PCE.
    • with P2P negotiation and minimal where IEs are exchanged, security is more complex as requires to secure this relation.
    • Ariton 15.4e includes several security fields that have a particular meaning. Would those fields have a different meaning when used in the context of 6TiSCH? What would be the impact of using or not using the MAC header.
    • Thomas the fields have the same meaning. What 6tisch defines is a key management protocol so certificates can be exchanged. This is a complement to the architecture and later the use of this security certificates is fully aligned with 15.4e std.

      Action: Ariton to send a mail to the ML asking clarification about whether IEEE802.15.4e fields have a different meaning when used in the context of 6TiSCH.

  • [09.11] RPL option status and update 6lo
    • Done at the previous minimal draft discussion.
  • RFC7322: RFC Style Guide
    • No time. Pushed to later call.
  • [09.13] Meeting ends
    • Next 6TiSCH call on 11/7, purpose is to go through material for IETF91 WG meeting.

Updated