I think it will depend on your v4<->v6 translation rules.
Peter Rathlev wrote:
On Mon, 2010-03-22 at 14:38 +0300, Tima Maryin wrote:
Why don't keep it ipv4?
If you run MPLS VPN netowork i don't see why you want something else
in core. MPLS don't care about ipv4/6 and PE just assign labels for
prefix/vpn and distribute it to other PEs.
For me it looks like double security here, if your core has no ipv6 i
can't be accessed (read: hacked) from outside ipv6 world/vpns.
It's no practical problem at all to keep the core IPv4. I was just
curious about why a pure IPv6 MPLS network wasn't possible (yet, on
Cisco boxen).
That leads me to another thing: If we _could_ make the core run IPv6,
and if we weren't using "no mpls ip propagate-ttl", how would a
traceroute from an IPv4-only machine work out? Would the core send
ICMPv6 Time Exceeded back to the host, which would then just discard
these unintelligible packets and show "*"s for that hop?
_______________________________________________
cisco-nsp mailing list [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/
