I'm after some opinions/advice on IPv6 rollout from an ISP perspective. First a general question: what's the general approach to assigning v6 to customers? Static vs dynamic, PD, WAN interface addressing? I've been reading that most seem to dish out a /64 for the CPE WAN interface plus a /56 (for ex...

Something isn't right there. It's normal to see the packet data in ICMP unreachable messages that generated it, but I can only see control plane packets, not data plane (the ping from CE1 to CE2).

Do you have multiple instances? You only need to create one aggregate per instance, so if Peer A, B and C are in the default instance, you only need to create the aggregate once.

Yes, on e0 on PE2 - this would be the point P2 removes the outer label, so you should still see the vpn label. If there's no label here at all, it explains why the traceroute times out at P1 and P2. Remember you're looking for the ICMP time exceeded packets originating from P1 and P2.

You only need a policy based route if you wanted only a subset of your LAN devices to use the second circuit, otherwise just use static routes for youtube.

When the TTL expires on either of the P routers, they have to build an ICMP reply to the source host. Unless the two P routers also have the customer VRF and routes, there is no way for them to build the ICMP messages back to the source. That, of course, defeats the object of running MPLS. The ICMP ...

I've just seen another thread saying RouterOS does not support label distribution via BGP. I suspect this could be the problem - P2 does not have VPN label knowledge, so when the TTL expires inside the MPLS core, P2 forwards each ICMP time exceeded message without a VPN label. With no VPN label, the...

Why use a route reflector? Just set up iBGP between PE1 and PE2. There's no need to run BGP on P routers with MPLS enabled. why wont the customer see the MPLS labels? Surely you don't want the customer to see your internal infrastructure? Don't forget that a LSP is one way. For the TTL to expire, th...

I'm not sure if this has been mentioned before, but I've noticed that setting an aggregate for an IPv6 prefix in BGP shows as origin code incomplete on ebgp neighbours. This is a problem because origin is chosen before MED, so the actual shortest path to the prefix is not being determined correctly....

For pppoe. Make all pppoe interfaces passiv in ospf interface configuration. Add pppoe /16 network to different area than backbone and use area range. to aggregate all /32 routes into on e/16 network. This is the proper way how to set up ospf in such cases. Sorry to drag up an old thread, but it's ...