Re: on the point of mobility & multihoming

[Masataka Ohta <mohta@necom830.hpcl.titech.ac.jp>]

john.loughney@nokia.com wrote:

> > From today's session, Dave & Geoff discussed multihoming vs. mobility.  In
> a working group I chair, we've had similar discussions and from a transport
> layer, maintaining a session during a multihoming change and a mobility change
> can be quite similar.

Not quite.

They have certain similarity that if we are going to solve mobility
issue in a way different from MIPv6, which is hopeless, minor details
of M6 design will be affected.

But, that's all.

Connection is a transport or application dependent concept, timing
of which is transport or applicaiton dependent.

OTOH, timing of mobility is governed by movemenet speed of mobile
hosts and service area of access routers.

> I second Dave's comment that trying to solve two
> things at once can be never-ending.

When I tried it last time, it took about half a year to design,
implement and install in the field. But as I think the
design was too much mobility centric, I have modified it,
part of which is the current 8plus8 proposal.

> Could we restrict ourselves to multihoming,
> but perhaps the authors of the proposals add a section on Mobility Considerations,
> so we don't run into incompatiple solutions?

As long as proposals keep the IP layer connectionless, which is
of course, even MIPv6 will work automatically (though with all
the defects) that there is not much to worry about.

Lear already wrote in his draft

: 2.1.2.1 Does your solution address mobility?
:
:    If so, how are rendezvous handled?  Can your solution handle both
:    locators changing at the same time?  If so, please explain. Should
:    it?  If not, how will your solution interact with MOBILEIP-V6 [3]
:    (MIPv6)

Is it enough?
						Masataka Ohta