Re: [6lowapp] Charter and transports

I would like to see the issue of transport added to the charter.

For smart energy, we plan to start with TCP, hoping that our relatively
small networks and ROLL don't violate the TCP congestion control timeout
threshold and that the code size of TCP are not too much for our small
devices.

This said, we would like a back-up plan.  We need to have UDP as that plan
but not using TCP removes many of the interesting applications that rely on
TCP.

I do think long term having transport in the charter for 6LowAPP/CoAp is a
necessary thing.

Don

-----Original Message-----
From: 6lowapp-bounces@ietf.org [mailto:6lowapp-bounces@ietf.org] On Behalf
Of Zach Shelby
Sent: Monday, November 02, 2009 10:07 PM
To: Jonathan Hui
Cc: 6lowapp@ietf.org
Subject: Re: [6lowapp] Charter and transports

Hi,

(New thread title)

So to me the answer seems to be that TCP will work for some  
configurations, but obviously not for all networks, and not for all  
requirements (e.g. multicast).

On Nov 3, 2009, at 5:15 , Jonathan Hui wrote:
>
>>> But are TCP's services/constraints
>>> appropriate for the application?  I can think of a variety of LLN
>>> applications that work just fine with TCP and others that could  
>>> stand
>>> to use something else to improve latency, message efficiency, non  
>>> p2p
>>> flows, multihoming, mobility,  etc.
>>
>> That's another question, yes. But I'd see building new transport  
>> layer functions for 6LOWAPPs as an activity that would need some  
>> very strong ties to the TSV area, if it wouldn't be hosted there in  
>> the first place.
>
> Agreed.  Though the line seems to be a bit blurred because some  
> suggest building transport-like mechanisms in whatever "app-level"  
> protocol we're working on.

In the Stockholm BarBof, Lars had a very good explanation on this one.  
We very well may start TSV area work on transport improvements - but  
that is a long-term effort and won't happen in time for CoAP. I do  
think we should aim at starting that work eventually, but let's get  
CoAP started first.

Obviously UDP is something we will need to support as a transport for  
CoAP, and I argue that we also need to allow the use of CoAP over TCP  
as this is useful for some applications. When using UDP, some simple  
mechanism for reliability (e.g. stop-and-wait) would of course be  
needed along with a transaction ID (see e.g. 6lowapp-frank-chopan).  
When doing this we should of course cooperate with TSV people.

So this brings us to the next question, what should we say in the CoAP  
charter about transports? Right now it says

"The protocol will operate over UDP...".

I would suggest it says something more like:

"The protocol will operate over UDP by default, and should define an  
alternative binding to TCP or another suitable reliable transport  
layer."

What do you think?

Zach

>
> --
> Jonathan Hui
>
> _______________________________________________
> 6lowapp mailing list
> 6lowapp@ietf.org
> https://www.ietf.org/mailman/listinfo/6lowapp

-- 
http://www.sensinode.com
http://zachshelby.org - My blog "On the Internet of Things"
Mobile: +358 40 7796297

Zach Shelby
Head of Research
Sensinode Ltd.
Kidekuja 2
88610 Vuokatti, FINLAND

This e-mail and all attached material are confidential and may contain  
legally privileged information. If you are not the intended recipient,  
please contact the sender and delete the e-mail from your system  
without producing, distributing or retaining copies thereof.

_______________________________________________
6lowapp mailing list
6lowapp@ietf.org
https://www.ietf.org/mailman/listinfo/6lowapp