Hi Ted,

On Wed, 22 Jun 2011 16:18:50 -0400
Ted Lemon <[email protected]> wrote:

> On Jun 22, 2011, at 7:41 AM, Mikael Abrahamsson wrote:
> > I agree, that's the deployment model I advocate for hostile scenarios. Use 
> > DHCPv6 for stateful addressing, advertise default GW via RA, don't 
> > advertise any on-link prefix, and make sure hosts can't L2 communicate at 
> > all with each other by means of enforcement in switches (or just separate 
> > them into different L2 domains).
> 
> Huh.   If I had a choice between RA and multicast, I think I'd choose 
> multicast.   When you have to utterly cripple your technology in order to 
> continue using a protocol, I think it's time to ask whether or not the 
> protocol you're using is the right protocol at all.
> 

You're right, with Ethernet being the wrong protocol. An Ethernet LAN's
nature is peer-to-peer or full mesh - every attached node can send to
any other individual node, many other nodes, or all other nodes. For SPs
who want to bill for, and possibly control/sanitise their inter-customer
traffic, this link layer peer-to-peer nature is a problem.
Hub-and-spoke or point-to-point reachability is what they want. If it
is possible to enforce a hub-or-spoke topology on an Ethernet LAN by
preventing the 1-to-many or 1-to-any capability, in effect making it an
NBMA link-layer, or creating a point-to-point topology via VLANs, then
Ethernet is the best choice because it is both cheap and ubiquitous. 

Regards,
Mark.
--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to