Hi Diego,
Yes, that's what I proposed.
The delay that
*starting from receiving the failure of 6P transaction, then forwarding to
SF and SF make the decision to start another 6P transcation*
depends on the position of current slot and next available slot for sending
6P request for Retry.
For ex
Correct Diego
And that's a desired effect so that a node gets a lot of bandwidth to complete
its join. You better complete a few at a time then have hundreds half way
through and none fully done.
We may actually reject a join if the queue is too long, exhausting=sleep well
and come back in 10
Pascal,
I was thinking of a worst-case condition: all neighbours are
willing to join. A limited pool would satisfy only a part of them, thus
putting the rest of the nodes in a join queue. Am I correct?
Regards,
Diego
2016-03-04 7:55 GMT-03:00 Pascal Thubert (pthub
Pascal,
Agreed. Then we should add a
new SHOULD item at the end of the
"4.2. Requirements for an SF"
section on the 6TiSCH Operation Sublayer (6top) draft
to include statistics.
Regards,
Diego
2016-03-04 8:21 GMT-03:00 Pascal Thubert (pthubert) :
> I’d support th
Pascal,
Acknowledge. Let's discuss tomorrow
on choice 2, as you mentioned.
Thank you.
Regards,
Diego
2016-03-04 8:06 GMT-03:00 Pascal Thubert (pthubert) :
> Hello Diego:
>
>
>
> You need a IANA section where you detail which registries you create and
> which you ad
Tengfei,
So, taking into account your proposal, the SF should configure
the Retry
Limit and handle the retries at SF layer, leaving 6P to deal only with
transactions:
either they are complete and successful or failed and rolled back (even in a
timeout case).
Retries shall not be
Hi Lijo
The usual way of doing this is to send it upon a change like parent or child
reboot, and upon reconnection. Then you send it again with an exponential Back
off.
The alternate is to be transactional and be able to resync, all of which is
easier with the bitmap than it is in the current
Hi Pascal,
This scheme sounds okay ,since it removes the need for 3rd transaction (ACK)
I feel the periodic sending of bit pattern is not required.
Thanks & Regards,
Lijo Thomas
From: 6tisch [mailto:6tisch-boun...@ietf.org] On Behalf Of Pascal Thubert
(pthubert)
Sent: 10 March 2016 13
Hello Qin
A chunk is like heap in malloc, a trick for allocation purpose by the parent to
get and defend many cells in one shot.
It is more link a “super bundle” than a chunk, since it is a bundle of cells
with some slots active and some not, depending on the bit mask that the parent
indicates
