Clone wiki

meetings / 160108_webex

Minutes, 08 January 2016 interim, 6TiSCH WG

Note: timestamps in PST.

Connection details

Taking notes (using Etherpad)

  1. Xavi Vilajosana
  2. Pascal Thubert
  3. Michael Richardson
  4. Tengfei Chang

Present (alphabetically)

  1. Thomas Watteyne
  2. Pascal Thubert
  3. C-Y Lee
  4. Diego Dujovne
  5. Jonathan Munoz
  6. Keoma Brun-Laguna
  7. Maria Rita Palattella
  8. Michael Richardson
  9. Michel Veillette
  10. Nicola Accettura
  11. Pat Kinney
  12. Qin Wang
  13. Rene Struik
  14. Tamer Elzayyat
  15. Tengfei Chang
  16. Xavi Vilajosana
  17. Zhuo Chen

Action Items

  1. Thomas to create an issue on 6top about ie format from TengFei's questions
  2. Maria-Rita to propose a scenario where chunks and bundles are being used
  3. Pascal to start a discussion on cell types (should there be chunk types?)
  4. Tengfei to ask what type of IES are (long or short) in 6top protocol.
  5. Xavi to take the abstract and intro and modify the current minimal draft.
  6. Chairs to talk to AD and agree on std track vs. bcp for minimal draft

Agenda

  • Administrivia [2min]
  • Approval agenda
  • Approval minutes last call
  • rechartering news
  • PlugTest [15min]
    • Venue info
    • Walk-through tests
    • related question about the 6lorh and 6top
  • 6LoRH
    • Figure13 in section 7
    • The meaning encapsulate address (source address?)
  • 6top-sublayer
    • Each 6P command require on response?
    • Typeof sixtop subIE:short or long?
    • Status of IANA ID for 6P
  • Minimal Draft [25min]
    • concluding last call on abstract / intro
    • impacts on main text
    • BCP or not BCP ?
    • Next steps [15min]
  • draft 6P protocol
  • draft SF0
  • AOB [1min]

Minutes

  • [07.05] Meeting starts

  • [07.07] Administrivia [2min]

    • Approval agenda
      • agenda is approved
    • Approval minutes last call
      • last call minutes are approved
    • rechartering news
      • Telechat on January 21
  • [07.??] PlugTest [15min]

    • Venue info
      • Paris, 2-4 Feb. INRIA. Near Gare de Lyon in Paris. Free Event. NDA. Registration 15 Jan deadline.
    • Sent test descriptions and tools to participants.
    • Walk-through tests by Xavi
      • test specs use 2 drafts:6top_sublayer and 6lo 6LoRH.
      • 13 tests, four of them warming. Not looking into the details of the packetsThese not as exhaustive as last plugtest.
      • Important piece is the 6top protocol test and then the 6LoRH
      • We'll verify addition, deletion, list and clear of cells through the protocol
      • We'll check the time out operation when the ack fails to come back
      • Will then check the RH3 and the RPI compressed by 6LoRH
      • This is a preliminary list, still going through in the coming weeks
    • Thomas: we are using the same tools as in Prague.
      • OpenMote can be sent early to participants to avoid basic failures
      • TengFei building golden image
    • TengFei: Going Through, golden should be ready in a few days [will this link to go mailing list?] I think so.
    • Thomas: did you get your questions answered
    • TengFei: I had questions. One was fig 13 and I got a reply from Pascal, fig will be removed.
      • The other thing I want to confirm is the IP in IP encapsulation last format: Is encapsulation address the same of the source address?
    • Pascal: Yes, this is the node that creates the IP in IP header, You have to provide it if it is not the root of the network but it can be elided if it is the root
    • Thomas: yes this is a substantial source of saving
    • TengFei: about 6Top protocol, is it mandatory to ack
    • Xavi: delete requires a response message indicating the success or failure of the operation.
    • TengFei: Do all the 6p commands require a response?
      • ACTION: thomas to add an issue in the bitbucket tracker to specify explicitly that all commands require a response
    • Tengfei to ask what type of IES are (long or short). What is the status of the IANA IDs in 6P?
    • Thomas: Choose numbers in the test description and later change if we get the IANA numbers.
    • Jonathan: The IE ID of them to use to update the dissector.
    • Thomas: use some defined numbers and write them in the test description.
    • Pascal: add a suggested value in the draft. If it makes sense IANA may actually grant that one
    • Pascal: Charter will be discussed the 21st of January. THen adoption call of the 6top sublayer (6P).
    • Xavi: will the 6lorh draft be split in two as said in the 6lo ML before the plugtest?
    • Pascal: it can be done.
    • Michael Richardson: If this happens the tools at ietf will link old and new version so this is not a big problem.
  • [07.32] Minimal Draft [25min]

    • concluding lass call on abstract / intro
    • Pascal: Last chance to do any modification. Conclude the call in the ML and indicate that the text is adopted.
    • Pat Kinney: typo in intro. 14.5 instead of 15.4 at page 2.
      • action> Xavi to take the abstract and intro and modify the current draft.
  • impacts on main text
    • Pascal: the coverage of protocols in minimal enough. Should it be more extensive?
    • Pascal: for me not. Architecture should be the general view. Minimal should collect the details for a basic configuration.
    • Pascal: architecture is informative reference to minimal which do not restrict in its publication.
  • BCP or not BCP ?

    • Pascal: BCP indicates how things are put together. It is not defining missing things but glueing.
    • Xavi: some other documents at IETF define a profile or glue several things and are considered standards. Xavi is fine with any approach as long as this is coherent with other documents
    • Xavi: as a second and more important consideration. We want this document to be adopted by the industry. If BCP constraints that adoption we want it to be considered in the standards track.
      • Action: PT and TW to talk to AD and take an executive decision.
  • [07.49] Next steps [15min]

    • draft SF0
    • Pascal: bundle is the set of cells between two neighbors. It is what is scheduled between A and B.
    • Pascal: Child should not care from which chunk it comes but it cares to which bundle it goes or from which bundle it is removes. We may have multiple bundles between nodes eg for tracks. We may also decide to have multiple bundles to represent ptiority.
    • Thomas: there may be different types of chunks to select from eg for priority.
    • Pascal: Well we may care for a type of chunk but not a particular chunk. Which exact chunk the cell comes from is the parent's cuisine. Today we have only one type for unicast cells with no risk of overlap but yes that should be discussed. eg we might have overlapping chunks which induce a risk of collision for use by statistic traffic of lower priority.
    • Maria-Rita: We should devise a simple scenario to position use of chunks vs. bundles
  • [08.06] AOB [1min]
    • Will Publish information on Plugtest location. INRIA Paris, near gare de Lyon. Register to get the information!
    • Need to start thinking about IETF 954. Duration and topics for the meeting. Will be an item for next interim
    • Do not forget to register to the PlugTest!
  • [08.07] Meeting ends

Updated