[6tisch] consideration on initial status in Figure 1, SF0 draft

2017-05-12 Thread Qin Wang 
Hi Diego,
Just below Figure 1, the text is as follows.
-1.  If 
REQUIREDCELLS<(SCHEDULEDCELLS-SF0THRESH), delete one or more       cells.2.  If 
(SCHEDULEDCELLS-SF0THRESH)<=REQUIREDCELLS<=SCHEDULEDCELLS, do       nothing.3.  
If SCHEDULEDCELLS<=REQUIREDCELLS, add one or more 
cells.-
I'm not very clear how it works. Here is my questions:1) what is the initial 
value of SCHEDULEDCELLS, SF0THRESH?2) what do you mean by " delete one or more 
cells" and "delete one or more cells"? Is there any formula to determine the 
number of cells to ADD/DELETE?3) After ADD or DELETE N cells, how to change 
SF0THRESH?
ThanksQin


___
6tisch mailing list
6tisch@ietf.org
https://www.ietf.org/mailman/listinfo/6tisch

[6tisch] Timeout implementation on 6P/SF0

2017-05-12 Thread Prof. Diego Dujovne 
Dear all,
During today's Webex call we discussed transaction
timeout implementation problems. As a result, the proposed
solution is that SF will include a TTL field with the timeout
value that the SF will wait to finish the transaction.
This will avoid inconsistencies if the transaction
is delayed due to retransmissions or intermediate queues.
As a consequence, the timeout value will be
implementation-specific.
Let me know if you agree in this item.
Regards,

 Diego



-- 
DIEGO DUJOVNE
Profesor Asociado
Escuela de Informática y Telecomunicaciones
Facultad de Ingeniería - Universidad Diego Portales - Chile
www.ingenieria.udp.cl
(56 2) 676 8125
___
6tisch mailing list
6tisch@ietf.org
https://www.ietf.org/mailman/listinfo/6tisch

[6tisch] Minutes, 12 May 2017 interim, 6TiSCH WG

2017-05-12 Thread Pascal Thubert (pthubert) 
Note: timestamps in PDT.

Connection details

  *   Date: 7-8am Pacific: 
http://www.worldtimebuddy.com/?qm=1&lid=100,12,5392171,1850147&h=100&date=2017-05-12&sln=14-15
  *   Webex link: 
https://cisco.webex.com/ciscosales/j.php?MTID=mcdbbe3a4e38d97d986b507ec12a1f9b1
  *   Webex Recording: 
https://cisco.webex.com/ciscosales/lsr.php?RCID=01514b03f5304ff6bbc0ae87d44b4bcf
 *   Recording password: Zj22kx7d
  *   Material link: 
https://bitbucket.org/6tisch/meetings/raw/master/170512_webex/slides_170512_webex.ppt

Present (previous meeting)

  1.  Thomas Watteyne
  2.  Pascal Thubert
  3.  A Paventhan
  4.  Diego Dujovne
  5.  Jonathan Munoz
  6.  Malisa Vucinic
  7.  Malisa Vucinic
  8.  Maria Rita Palattella
  9.  Michael Richardson
  10. Remy Leone
  11. Sumankumar Panchal
  12. Tengfei Chang
  13. Yasuyuki Tanaka

Action Items (previous meeting)

  *   Thomas to contact Carsten
  *   Thomas to contact Maria Rita and Miguel and coordinate PlugTest
  *   Xavi and Qin to review SF0
  *   Jonathan to review 6P after Charlie's comments are addressed
  *   Xavi and Qin to address Charlie's comments
  *   Xavi to bootrap research liaison
  *   Diego to write an email and open a ticket to track the issues between SF 
and 6P

Taking notes (using Etherpad)

  1.  Pascal Thubert
  2.  Thomas Watteyne

Action Items

  *   Diego to start a thread on time out computation between SF0 and 6P. 
Whether ASN of the time out should be indicated in the packet.

Agenda

*Administrivia [ 7min]

 *   Agenda bashing
 *   Approval minutes from last meeting
 *   Addressing todo's from last time
 *   News from PlugTest

*6P finalization (Thomas, Qin) [20min]

  *   SF0 finalization (Diego) [20min]
  *   Update on security (Michael/Malisa) [ 5min]
  *   AOB [ 3min]

Minutes

  *   [07:05]Administrivia [ 7min]
 *   Agenda bashing
 *   Approval minutes from last meeting => Approved
 *   Minutes of last meeting; this agenda approval
*   [Michael] "MR" in the todos refers to Maria Rita, not Michael 
Richardson
*   no other comments
 *   Addressing todo from last time
 *   News from PlugTest

*[7:12] 6P finalization (Qin) [20min]

 *   Addressing Charlie's comments, most of them accepted
 *   Qin: Charlie suggests a terminology section. We have a terminology 
document.
 *   Tech comments are addressed one by one, list below.
 *   TW: metadata is not always the same but is limited to 2 bytes
 *   Qin: How to express in the draft. We say same usage, gives impression 
that the meta data is the same.
 *   Thomas: we may say same usage but not same content
 *   Qin: OK
 *   Qin: On 6P vs. SF0, some considerations being potentially impracticla. 
need to chat with SF0 people to address feasibility.
 *   Thomas: about aggregating SFs, well not supported
 *   Thomas: Registry does not exist, lost cause to start one now.
 *   Diego: there was also the requiremtn to define time out but I think it 
does not belong to SF0. Should be in 6P. SF0 has no tools to defined time outs.
 *   Thomas: Why
 *   Diego: we do not know how long to wait. Would that be the longest time 
possible? Impractical. Each transaction may be different. SF0 can handle the 
time out but not compute the value of the timer.
 *   Thomas: The time out depend on the schedule
 *   Qin: different SF will have a same or different values?
 *   Thomas: the SF may know that it needs an adiitional delay.
 *   Diego: SF0 has no control of the duration
 *   Pascal: 6P could return a duration when SF0 sends a packet and SF0 
includes that in its time out.
 *   Pascal: SF on node A adds in the request the ASN of the time out
 *   Thomas yes we need a thread on that

*[7:42] SF0 finalization (Diego) [20min]

 *   Diego: there were discussions on terms and functions (too fast for 
minute taker, see recording).
 *   there are 2 proposed algorithms; very different from one another. The 
changes only impact the first. One passes data to the other, that was clarified.
 *   Qin: Initial status of SF0? What time are overprovisionned cells added?
 *   initial status is 0 cells. We are only having overprovisionned that 
were used cells in previous instance (next time).
 *   Thomas: figure 1 on page 5

*[7:52] Update on security (Michael/Malisa) [ 5min]

 *   Michael has no update at this time

*[7:57] AOB

___
6tisch mailing list
6tisch@ietf.org
https://www.ietf.org/mailman/listinfo/6tisch