Re: [Iot-directorate] Question on IETF IOT-devices scope

[Toerless Eckert <[e-mail]>]

Samita:

To answer your question:

a) Like others on this thread, i think we never limited IoT to "constrained" devices.
We just did a lot more work for them than for other (T)hings (like cars and their
networks. I have not followed SUIT, but i could imagine that SUIT and BRSKI work
will reach far beyond constraint Io(T)hings already.

b) I already commented on earlier version(s) of Ted drafts, that i wonder
about the applicability of the work transcending the specific use-cases Ted
has in mind into other (T)hings. I just have not seen good feedback from
folks that could comment from more experience from other use cases, such as
in industrial (T)hings.

For example, from my limited understanding, there are a lot of networks
in industrial use where only IPv4 is used because they like the idea of
being able to achieve really very simple (T)hing software stack by using
fixed addresses, such as 10.0.0.x for the components of some machinery
connected by a simple ethernet and a "gateway" switch. These setups are
then also why you find a lot of static NAT functionality in indistrial
ethernet switches - to connect such stub networks to the next layer of networking.

With ULA in IPv6 we do not even allow for this "generic" addressing model
anymore because you are supposed to allocate a random ULA prefix, and NAT
is looked upon with despise by many IPv6 purists, whereas IPv6/IPv6 NAT6
did evolve though many use cases that evolved from the NAT mindset from that
IPv4 history IMHO. I remember the heated debate about ading IPv6 NAT to
OpenWRT several years back where purists said it shouldn't be done and
i think it exists now in OpenWRT but of course the debate goes on.

Yada yada: I feel that for broader use across various IoT use cases it would
certainly help to better understand addressing desires in more IoT network
types especially industrial. Don't be surprised if there is a lot more
interest in better IPv4 support and all types of NAT as well as "generic"
hard-coded addresses. And i am not making judgement calls 

Cheers
    Toerless

On Mon, Mar 15, 2021 at 05:55:49PM -0400, Chakrabarti, Samita wrote:
> Hello All:
> 
> In pre-IETF IOT-DIRECTORATE meeting, Ted Lemon has pointed to two documents
> regarding connecting the Thread Mesh networks to the traditional IP
> Infrastructure:
> 1) (https://datatracker.ietf.org/doc/draft-lemon-stub-networks-ps/)  -
> Problem Statement
> 2) https://datatracker.ietf.org/doc/draft-lemon-stub-networks/         - A
> possible solution
> 
> IoT-Directorate co-chairs and the responsible ADs would like to discuss if
> it makes sense to extend the IOT-Directorate charter towards less
> constrained devices such as some of the home devices considered as 'things'.
> 
> The charter of the directorate says : (
> https://datatracker.ietf.org/group/iotdir/about/)
> 
> "The Internet of Things (IoT) refers to devices, that are often constrained
> in communication and computation capabilities, now becoming more commonly
> connected to the Internet, and various services that are built on top of
> the capabilities these devices jointly provide. It is expected that this
> development will usher in more machine-to-machine communication using the
> Internet with no human user actively involved."
> 
> Should IoT-Directorate members consider Thread Mesh stub networks as one of
> topics under IOT-DIR scope of discussion or part of IoT work?
> 
> Thanks,
> 
> -Samita

> -- 
> Iot-directorate mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/iot-directorate


-- 
---
[e-mail]