I agree with Brian Carpenter,
"We expect millions of those during v6/v4 coexistence."
So back to my original question, which apparently none of
the IPv6-Leaders liked:
-- if we are doing tunnels which follow a logical
topology rather than a physical one,
-- why don't we have support for multihoming to
different logical topologies
We should. But multihoming is still a hard problem and we are
still working on it in IPNGWG.
Multihoming is not a hard problem.
None the less, IPNG WG have been working on it in vain only to make
protocols complex, because it is working for a wrong direction.
Multihoming does not scale and is not reliable, if it is offered by
intelligent intermediate systems.
IPv4 multihoming does not scale and is not reliable, because it is
offered by intelligent routing systems.
Multihoming scales and is reliable and easy, if it is offered end to
end by intelligent end systems, transport/application protocols on
which directly handles all the multiple IP addresses from DNS
However, partly because the committee has worked too much to a wrong
direction to offer multihoming by intelligent intermediate systems
and partly because the committee decided to minimize code modification,
it is likely that they can't accept the fact of internetworking that
only end-to-end approach scales and will keep working for the current
This memo describes the architecture of end to end multihoming.
End to end multihoming does not burden routing system for
multihoming. That is, even extensive use of end to end multihoming
does not increase the number of entries in a global routing table.
Traditionally with IPv4, multihoming capability is offered by an
intelligent routing system, which, as is always the case with
violating the end to end principle, lacks scalability on a global
routing table size and robustness against link failures.
On the other hand, with end to end multihoming, multihoming is
supported by transport (TCP) or application layer (UDP etc.) of end
systems and does not introduce any problem in the network and works
as long as there is some connectivity between the end systems.
Because end to end multihoming is performed in end systems, the
architecture needs no routing protocol changes. Instead, APIs and
applications must be modified to some extent.
for more details.