Re: IPv6 multihoming scenarios I-D

[Iljitsch van Beijnum <iljitsch@muada.com>]

On 12-jul-04, at 20:18, JORDI PALET MARTINEZ wrote:

> As consequence of some comments with the WG co-chairs in the recent  
> interim meeting (see the minutes also), we submitted this morning this  
> document (for the inpatients):

What about the outpatients?  :-)

> http://www.consulintel.euro6ix.org/ietf/draft-palet-multi6-scenarios 
> -00.txt

The first part is very good. In fact, much of this should have been in  
the "how it's done today" draft. Maybe it would even be good to merge  
these two drafts, as the text overlaps significantly even if the  
intention behind the documents doesn't.

One thing that could use a little more text is the situation where a  
network connects to the commodity internet through one or more ISPs,  
but also to a specialized network such as an NREN. In this case the  
policy that certain traffic should always flow over the specialized  
network, or that certain traffic never flows over the specialized  
network is important and must be enforceable.

> There is still some work to be done, and we will like to get some  
> inputs specially from different experts:
> - 3GPP
> - Ad-hoc
> - Mesh
> - Mobility
> - Content Delivery Networks
> - Internet Data Centers
> - Other special applications or protocols ?

I wasn't impressed by this part. Sure, an UMTS network can be  
multihomed, and a mobile device can be multihomed. But does this mean  
we have to talk about 3GPP (what does this stand for anyway?)  
multihoming? Better remove this part and stick with what's important.

Iljitsch