On Wed, 29 Aug 2007 22:08:26 -0400 (EDT) [EMAIL PROTECTED] wrote: > > Brian D, > > <snip> > > These days, for example, on an IPv6-*only* Cisco 7600 with 3BXL, will > only hold a theoretical maximum of 119,000 prefixes. Bigger iron may > support double that, but again, only as IPv6-ONLY. When you add IPv4 > onto the same platform, the huge IPv4 DFZ eats most of the capacity > for IPv6 prefixes. Most providers will need to have dual-stack for > a considerable period of time (3-5 years minimum, IMHO.)
I'm inclined to believe that dual-stack provider networks are going to be relatively rare, and may not exist at all. I think it'll either be IPv6 over MPLS (in one way or another e.g. 6PE or native IPv6 over MPLS), or Softwires, where IPv6 is encapsulated inside IPv4 (or vise-versa) at the edges, and tunnelled across a native IPv4 or IPv6 backbone to the egress device. I think softwires is also likely to be the better alternative for enterprise networks as well, although the parallel introduction of IPv6 and BGP may initially be a bit intimidating for them. Assuming a Softwires IPv4 over a native IPv6 network, one method of avoiding having to implement BGP could be to encode IPv4 prefix information below a reserved IPv6 prefix, and then have the normal IPv6 IGP carry that information between the egress and ingress nodes. Both of these techniques push the "heavy lifting" of matching a destination layer 3 addresses against many and specific prefixes to the edge of the network. Once the problem can be pushed to the edge of the network, it can typically be scaled more easily. The TCAM limit in your high end internal devices then becomes a limit on how many forwarding devices you have in your network, not how many prefixes the network can forward for. Regards, Mark. -------------------------------------------------------------------- IETF IPv6 working group mailing list ipv6@ietf.org Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------