Jun-ichiro itojun Hagino wrote:
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
WIDE and two of the biggest japanese ISPs are already dual-stacked,
and if my guess is right it has been so for more than 5 years.
it is not economical to purchase separate trans-pafici circuit.
Itojun,
I believe he was referring to the *method* of supporting both IPv4 and
IPv6, and arguing that core routers will tend to be addressed with
either IPv4 or IPv6, but not both. Instead, support for the second
protocol will tend to be provided by edge routers tunneling across the
core (using 6PE, softwires, or another method), rather than by native IP
forwarding by core routers. These solutions don't require separate
circuits for IPv6 and IPv4 traffic.
Are you familiar enough with the internal addressing of those Japanese
ISPs to say whether they address their core routers with both IPv6 and IPv4?
Another consideration is how big a given network's "core" is. In the
case of a geographically constrained network, it may be quite small, and
a dual-stack core may make more sense than MPLS. For large global
networks, however, it appears to me that tunneling across the core is
becoming the preferred mode of operation.
-Scott
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------