Re: Question re HIP dependency [Re: about Wedgelayer 3.5 / Fat IP approaches]

[Erik Nordmark <Erik.Nordmark@sun.com>]

> > Question to the WG: given the current state of HIP, do we
> > want to consider dependency on HIP as
> 
> > a) acceptable
> > b) unacceptable?
> 
> Unless I'm mistaken HIP is still work in progress so that doesn't help. 
> Apart from that I have two concerns: the identifier space is flat, 
> which can create problems such as with referrals, and HIP as it is 
> today requires the use of IPsec, which adds processing and bandwidth 
> overhead whether or not this is desired or appropriate for a certain 
> communication session.
> 
> This leads me to believe that using HIP as-is isn't a very good idea 
> for multi6, but it's certainly conceivable that a good multihoming 
> solution could be created based largely on HIP.

Agree fully.

HIP does provide worked out pieces for some of the components that might be
needed, hence I certainly look at it very closely.
But for the two technical reasons Iljitsch mentions, I don't think
HIP as currently specified fits what we need. I want a multihoming solution
that can allow existing applications to do referrals, and I don't
want to require that every packet be encrypted.

  Erik


   Erik