Re: [core] Time Scale Option

Hi Klaus,

Yes, ACK_TIMEOUT and MAX_LATENCY are known from the LPWAN device and the
value carried in the option result of this computation.

Laurent

On Tue, May 29, 2018 at 11:37 AM Klaus Hartke <hartke@projectcool.de> wrote:

> Wouldn't ACK_TIMEOUT, MAX_LATENCY, etc. need to be adjusted, since
> EXCHANGE_LIFETIME is just a calculated value?
>
> Klaus
>
>
> On 29 May 2018 at 11:01, Laurent Toutain
> <laurent.toutain@imt-atlantique.fr> wrote:
> > Hi,
> >
> >
> >
> > We have published few month ago a draft defining a new CoAP option.
> >
> >
> > https://datatracker.ietf.org/doc/draft-toutain-core-time-scale/
> >
> >
> > The main idea is to inform a server when devices will not have the same
> > reaction time. Current CoAP RFC defines a default 5 minutes window during
> > which any message will be consider as a retransmission.  This delay is
> too
> > small for LPWAN network, where small bandwidth, duty cycles are
> introducing
> > larger delays. One solution is to change the default window value, but
> this
> > period will be too long for “regular” devices and force the server to
> > memorize more message ID.
> >
> >
> >
> > That’s why we propose the Time Scale option, the client informs the
> server
> > of the window period where messages will be considered as retransmitted.
> The
> > server will be informed for this device and will adapt its timers.
> >
> >
> >
> > We would like to know, as we discussed during London IETF, what is the
> > position of the core wg on this option.
> >
> >
> >
> > Ana and Laurent
> >
> >
> >
> > _______________________________________________
> > core mailing list
> > core@ietf.org
> > https://www.ietf.org/mailman/listinfo/core
> >
>