Re: Minutes / Notes

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

On zaterdag, jul 19, 2003, at 19:15 Europe/Amsterdam, J. Noel Chiappa 
wrote:

> I don't think it's realistic to propose the addition of two separate
> mechanisms to do multi-homing, with slightly different capability sets 
> - one
> able to keep conections open, and one not (I assume that this latter 
> is done
> by picking one of N addresses, with no capability to change it).
I agree that having two or more mechanisms is undesirable in the long 
run.

On the other hand, solving the source address / ingress filtering issue 
(which we almost certainly need to do for loc/id anyway) would make it 
possible to achieve many multihoming benefits by only slightly 
modifiying applications, i.e., let them cycle through all possible 
addresses. This is of course not nearly as good as a full solution, but 
it is much better than what we have now.

I'm not a big fan of bringing mobility in the picture, though. I 
haven't seen any figures on the implementation status of MIPv6 in our 
collective favorite OSes, but it hardly seems omnipresent. When also 
considering the complexity of MIPv6 and the fact that MIP assumptions 
fail badly for multihoming I don't see much value in a MIP approach to 
multihoming.

(If I could I would even like to go further by scrapping MIPv6 
wholesale and starting fresh with loc/id for both multihoming and 
mobility.)