I have a requirement for a PPP Client who is only utilising IPv6 to be able to ping an IPv4 only end point.... Let me show you the flow of traffic:
The actual L2TP tunnel is initiated between two IPv4 Peers and should have no influence on the contents of the tunnel.
PPP Client (S0/0 - IPv6) --> S0/0 (PPP)LAC --> ge-1/3/1(PPP) LNS --> (ISIS IGP) MX240 Core --> Transit (IPv4 or 6)
The test can be to ping from the PPP Client to the far end (Transit IPv4) and see if we get a response. In the real world the PPP Client will be our customer who wants to traverse our ISP network and maybe contact a pure IPv4 interface on the internet, or a server somewhere.... this is what we need to replicate and have working....
What is the best way to achieve this and how?
Nat64 at the client ---- Fa0/0 IPv4/IPv6 (Customer facing CPE interface) - S0/0 IPv6 (LAC facing interface) on the client PPP...
Actually, I'm assuming that the NAT will be required on the PPP Client for the translation between the customer interface and the PPP Interface.
If the exit to the transit is out of an MX240 at the other end, is this where the translation back needs to take place?
Okay. I will try NAT64 at the PPP Client..... on the MX240 Core, I will have to look at how that conversion back will happen. As that is going to connect directly with the transits then 2 things need to occur there:
1: Whatever NAT took place needs to be converted back (as only IPv6 will be on the PPP Client interface).
2: BGP policies will be in place here.
Any help on the best way to achieve that would be great....