Re: [6lo] working group last call (wg lc) on https://datatracker.ietf.org/doc/draft-ietf-6lo-deadline-time/

["Georgios Z. Papadopoulos" <[e-mail]>]

Hello Lijo,

Thank you so much for your detailed comments. I appreciate it very much.
I am happy with your response, I just have one last clarification point, see below:


> On Jul 24, 2018, at 09:38, Lijo Thomas <[e-mail]> wrote:
> 
> Dear Georgios,
>  
> Thanks for your valuable suggestions and we really appreciate for taking your valuable time for the review .
>  
> Please find our comments inline below marked as (*** [LT]) 
>  
> We will be happy to receive your further inputs !!!
>  
>  
> Thanks & Regards, <>
> Lijo Thomas 
> 
>  
> From: <> 6lo [mailto:[e-mail] <mailto:[e-mail]>] On Behalf Of Georgios Z. Papadopoulos
> Sent: 17 July 2018 21:40
> To: [e-mail] <mailto:[e-mail]>
> Cc: [e-mail] <mailto:[e-mail]>; [e-mail] <mailto:[e-mail]>; Malati Hegde; Samita Chakrabarti; Gabriel Montenegro; lo; Charlie Perkins; [e-mail] <mailto:[e-mail]>
> Subject: Re: [6lo] working group last call (wg lc) on https://datatracker.ietf.org/doc/draft-ietf-6lo-deadline-time/ <https://datatracker.ietf.org/doc/draft-ietf-6lo-deadline-time/>
>  
> Dear Lijo and co-authors,
>  
> I went through the draft, please find my comments below:
> — — 
>  
> High level comments:
> */ [GP] The draft defines the Deadline Time (DT), but it is not clear to me how the arrival of the datagram within this pre-defined DT period is guaranteed?
> Indeed, the draft provides the necessary DT information, however, the only action I could observe is the delay-sensitive datagram to be dropped if the indicated DT is elapsed.
>  
>  
> *** [LT] Yes, the Deadline Time (DT) specifies the maximum allowable delay
> before which the packet should be delivered to the destination. The proposed
> draft provides a mechanism for transporting the DT information. By incorporating
> deadline based scheduling/routing mechanisms within the intermediate nodes
> using DT, one could guarantee deterministic behavior in terms of delay. 


[GP] Would you agree that this draft do not guarantees deterministic behavior in terms of delay, but it provides
the information of maximum allowable delay for a packet to be delivered to the destination?

To be more precise, for instance, lets us consider the following multi-hop network A—> B —> C.
According this draft, it will required 2 timeslots (or 20ms) for a packet to be delivered at the DODAG Root C.
However, if there is an external interference from A to B, then A may need to retransmit multiple times
in order the datagram to be received by B. Then there are two options according to the draft:
a) the datagram is dropped, to reduce the traffic, energy consumption.
b) the datagram is delivered even if the deadline time is crossed, i.e., as you said in your e-mail “in some scenarios where the intention is also to know the total delay experienced by the packets in a network”

In both bases, a and b, there is no guarantee that the datagram will be delivered in predefined time, i.e., in deterministic behavior. 

— — 
Thank you so much,
Georgios

____________________________________

Georgios Z. Papadopoulos, Ph.D.
Associate Professor, IMT Atlantique, Rennes

web: 	 www.georgiospapadopoulos.com <http://www.georgiospapadopoulos.com/>
twitter: 	@gzpapadopoulos <https://twitter.com/gzpapadopoulos?ref_src=twsrc%5Etfw&ref_url=http://georgiospapadopoulos.com/>
____________________________________