RE: Automatic tunnels

["Christian Huitema" <huitema@windows.microsoft.com>]

> We also need to understand if and how a home ISP can deploy gradually
> IPv6
> for its customer. Does providing a 6to4 relay helps? Does providing a
> tunnel broker
> in in core/access networks help? Does this makes it easier to go
> gradually to native?

I guess we agree that the main priority for ISP ought to be, "provide
native IPv6 services". The question then is what should an ISP do when
it wishes to provide IPv6 but is not quite ready. I would expect the ISP
design team to tell us more about the constraints and the
practicalities, but there are clearly three possibilities: do nothing,
provide support for automatic tunneling, or provide configured tunnels.
The support for automatic tunneling is simple: just provide a
6to4-to-native service and advertise a route to 192.88.99.1; to avoid
abuse, make sure that 192.88.99.1 is only reachable by subscribers of
this ISP. The support for configured tunnels is somewhat more complex:
if the ISP really want to provide additional benefits compare to
automatic tunnels, it needs to do some form of customer management to
ensure that customers get stable prefixes, that they use the tunnel end
point closest to home, etc. Frankly, I don't know whether the additional
complexity is worth the cost.

> The issue with with an ISP deploying an internal 6to4-to-native relay
> is that
> it will help its customers who are using 6to4 to send packets to
native
> IPv6
> destination, but won't do much to make sure that the same packets will
> return any quicker. As 6to4 to native and back is fundamentally
> asymmetric routing,
> somebody, preferably close to the native IPv6 node has to have a route
> for 2002::/16.

Deploying a "native to 6to4" relay is actually very easy -- its is a
matter of turning on 6to4 relaying on a dual-stack router, or even on a
dual-stack host. I would expect that these relays will be plentiful.

> As one can not realistically expect to have every single IPv6 network
> maintaining
> a local native-to-6to4 gateway, we would have to rely on public,
> open native-to-6to4 relays to exist and advertise 2002::/16 in the
DFZ.

We should precisely aim for having at least one native to 6to4 router in
every dual stack ISP. These routers are very easy to control: just don't
propagate 2002::/16 in BGP. They provide an immediate added value to the
local IPv6 users.

> So far, I have identified only a few of those public native-to-6to4
> relays.
> My understanding is that, expect from the willingness to promote the
> technology,
> there is actually little incentive to operate such a relay, as it will
> just attract transit traffic
> that has nothing to do with the ISP customers.

I believe it is a whole lot easier to provide "private" relays, for
local use only, rather than public relays, for the whole Internet.
However, backbone ISP have a built-in incentive to provide these relays:
the relays create traffic, and that traffic ends up being billed to the
second tier ISP who send packets to the backbone.

-- Christian Huitema