Webex CALL, May 9th, 2014
- Webex Link
- Live etherpad for minutes
- Topic: 6TiSCH Weekly
- Time: 8:00 am, Pacific Daylight Time (San Francisco, GMT-07:00)
- Meeting Number: 206 802 913
- Meeting Password: sixtus
- CCM: +14085256800x206802913
- Raghu to start a thread on ack mechanism for the 6top blocks. Pat to describe the ack vs. Enhanced ack.
- Pouria to start a thread on which code is reused for managing the block retries.
- Pascal to doodle a plugfest time for the beginning of the plugfest on Sunday before IETF
- [10min] Administrivia
- Approval agenda
- Approval minutes last call
- IETF 90 in Toronto: Registration opened;
- PlugFest ? bits and bytes ?
- draft-ietf-6tisch-coap published
- [10min] Status on 6top discussions
- [20min] Slotframe structure in Minimal draft
- [20min] Terminology discussion
- [2min] Next meetings
- [1min] AOB
Meeting starts at 08.05 PST
Taking notes (using Etherpad)
- Xavi Vilajosana
- Pascal Thubert
- Bryan Mc Laughlin
- James Pylakutty
- Kazushi Muraoka
- Kuor Hsin Chang
- Maria Rita Palattella
- Michael Richardson
- Nicola Accettura
- Pascal Thubert
- Pat Kinney
- Pouria Zand
- Qin Wang
- Raghuram Sudhaakar
- Rouhollah Nabati
- Xavi Vilajosana
- [8:05] Meeting starts
Agenda is approved. Last Call minutes are approved
[8:08] Update IETF 90 -registration is open.
[8:09]Coap Draft has been accepted as a WG document.
Milestones: WG documents adopted.
Bits and Bytes -to expose and show product demos -the subject of B&B is IoT
Plugfest As alternative Possible days: 1.Sunday Afternoon, 1.Monday Evening 1.Wed Evening 1.Friday
ACTION to contact other IETF groups to inform about plugfest.
Michael: Friday afternoon is no good since network is rapidly shut down. Late hours not good either people are tired.
Suggestion to do it on Sunday Morning.
ACTION: DOODLE at the ML.
Xavi: discussions on block option to fragment packet presented by Qin Wang, who presented what the wrapper IE looks like. The we discussed overhead and alternatives (fragments at app level. Result is that the block option already has everything we need, so better reuse. Overhead appears to be 17 out of 64 (see slides) Win also discussed whether the URI path is needed since it is supposed to be in each block. Unclear so far.
Pouria: Do we need to have the code for retransmission in 6Top.
Xavi: at MAC layer we do not know if parsing works so we need all the code in 6top
Raghu: I agree, should not use the MAC reliability mechanism at 6Top. Best to separate the mechanisms, they do not work exactly the same some not the exact same code. Code for IE part ans well as comm with PCE is the same to it is still advantageous.
Xaxi: 6top acks will be packets not mac layer ACK. Can we you the wrapper IE?
Raghu : Yes we need to explore that offline.
Pat Kinney: if we have a6Top ack on top of the ack we add a significant overhead
Raghu: agree but the concern is how the reassembly happens
Pat: let us off line the discussion, maybe we can change the MAC
Pascal: if the error is an exception then have an exception message not an additional ack every time.
Pat: there is the legacy ACK and the enhanced ack. The e-ack checks for the MIC on top of CRC, let us offline
Kuor Hsin: merging the acks is being discussed
Xavi: URI-host is definitely an option and we can elide it
Xavi : proposal to use OTF?
Pascal: This does not allow offseting multiple minimal networks in sytnc with a backbone interfereing in a same room. let us continue on the thread.
Xavi presents the minimal schedule. Problem is very low bandwidth and sync is all at teh same time. For PB2, we can space the slots uniformely to the sync interval is reduced to 300ms instead of 1.3s
difference between re-join and reboot. Pascal proposes to have an initial phase
Discuss terminology next meeting
- No other Business; adjourning.
[8:59] Meeting ends