I think the draft is a bit optimistic with respect to its claim of satisfying 
homenet requirements. For example, there is a very limited set of multi-homing 
topologies where the internal routers don't find themselves asking, "Which way 
is up?". Additionally, without source routing, I don't see how ingress 
filtering can be avoided. Perhaps multi-homing should be removed. 
Thanks,
Acee 

On Mar 13, 2013, at 6:09 PM, Tim Chown wrote:

> 
> On 26 Feb 2013, at 15:46, Chris Grundemann <[email protected]> wrote:
> 
>> Hail Homenet,
>> 
>> FYI - We have updated our "HIPnet" draft. This -01 is a minor update,
>> which includes almost exclusively editorial fixes. We have taken note of
>> all feedback received thus far, both on and off list, and will work to
>> incorporate that feedback in further revisions as it makes sense.
> 
> Hi Chris,
> 
> It's good to see a 'lightweight' alternative being put forward, with a demo 
> here.
> 
> From the arch text point of view I agree it's largely compatible with -07, 
> assuming the text about 'routing protocol' were rewritten to 'routing 
> functionality', since you install routes automatically up/down.
> 
> The two points that are partly met are the prefix utilisation efficiency 
> (since it's hierarchical DHCP-PD) and whether arbitrary topologies are 
> supported (given the use of ND in a 'spanning tree' like approach). But it 
> does tick a lot of boxes, and uses existing protocols.  Topics such as realms 
> and borders, and naming/SD are orthogonal to how the prefixes and routing is 
> configured.
> 
> The interesting question for the WG is how such an approach could 
> interoperate with (say) a homenet solution using zOSPF. Are the homenet 
> solutions 'all or nothing', or is there something we could do now to improve 
> the chances of future interoperability if hipnet is deployed now?
> 
> Tim
> _______________________________________________
> homenet mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/homenet

_______________________________________________
homenet mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to