Thanks Eliot
Would be useful i think to have the fairhair group consider some options:
- If devices use some "mesh" like topology, they easily could become
transit node, then RPL is a very lightweight routing option.
- Even if device is a leaf, participating in RPL would be comparable
lightweight to eg: DHCP or other protocols (hopefully some expecrts
can chime in to refute or support that statement ;-). And RPL would
allow to simplify learning parameters and addressing.
- Instead of SLAAC/DHCP, it might be quite interesting to consider
an addressing model as we have it in ANIMA: Where the address is effectively
configured during bootstrap (in ANIMA case into the Certificate).
- With this type of statically provisioned IPv6 address on device and
RPL, it is extremely simple to keep track of the device in a database
(no dynamic remapping of IP address to device required), and on
the RPL root (which could be co-located with a controller, you would
also see as RPL routes all devices present in the network by their
address.
These IMHO are the most attractive options from the full anima model
(beyond bootstrap) for lightweight IoT devices - RPL routing/bootstrap
provisioned static IPv6 addresses. Eliminates IMHO some layers of complexity.
Cheers
Toerless
On Tue, Aug 16, 2016 at 10:17:37AM +0200, Eliot Lear wrote:
>
>
> On 8/16/16 10:06 AM, Toerless Eckert wrote:
> > Thanks, Eliott
> >
> > Quick questions:
> >
> > a) Would those Fairhair devices ever be transit nodes ("routers") in the
> > network itself ? Or even when only being leafs of the topology,
> > would they intend to run RPL ?
>
> TBD, I think.
>
> >
> > b) Any specific way how these devices get their IP (IPv6 ?) addresses ?
>
> There will be multiple models in play, based on the operational
> environment. In some cases, 802.15.4 networks will be used, in others,
> wired, in some cases WiFi. Likely devices will start out with IPv4 and
> require DHCP, and then in IPv6 both SLAAC and DHCP will be required.
>
> But this is my personal view.
>
> Eliot
>
>
--
---
Toerless Eckert, eck...@cisco.com
_______________________________________________
Anima mailing list
Anima@ietf.org
https://www.ietf.org/mailman/listinfo/anima
