[e2e] node addresses vs. interface addresses
touch at ISI.EDU
Wed Jul 31 23:18:09 PDT 2002
J. Noel Chiappa wrote:
> > From: "Tim Moors" <t.moors at unsw.edu.au>
> > Perlman explains how "internal network numbers" in IPX helped avoid
> > sub-optimal routes to multihomed destinations
> Except that this is a bad idea in a very large network. In short, it's
> putting the entire cost of multi-homing (one rarely gets a benefit without
> some cost somewhere, neh?) into the routing, where the cost is paid by
> everyone (effectively - long arcane routing point elided) across the network,
> to maintain a route to that destination. This is fine in a small network,
> where the total incremental cost is small, but not fine in a big one.
Except again that this can be hidden from the rest of the network using
tunnels. A host can have a single endpoint address - on an "internal"
virtual interface, used through a set of tunnels, each using outer
addresses based on one of multiple real interfaces.
This allows the connection to switch between interfaces without burying
routing protocols into the transport protocol. It also allows the
routing protocol to be hidden except at the multihomed endpoints - i.e.,
it's used only during and for the tunnels. That's a somewhat localized
use of routing protocols, but it can isolate them from leaking out into
the rest of the net.
IMO, it's better than wedging routing (liveness, path detection, and
failover) inside the transport (ala SCTP).
> PS: If anyone cares about an architectural perspective on multihoming, I
> have some notes on it here:
Mine, FWIW, were implemented as:
"Dynamic Host Routing for Production Use of Developmental Networks"
J. Touch and T. Faber, Proc. ICNP '97, Atlanta, Oct. 1997, pp. 285-292.
They also form the basis of our variant of overlay networks (the
X-Bone), an architecture which we are currently developing (see also
More information about the end2end-interest