The MPLS WG Archive[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index][Thread Index][Author Index][Subject Index] [mpls] MPLS over L2TPv3 as WG item
Mark, Thanks for your response, a few replies inline to provide clarification. > > If an operator already has L2TP deployed, why not just > > continue to use L2TP VPNs? Why go to the trouble of > > upgrading the edge to support MPLS and leave L2TP > > in the core? > > L2TP(v3) is an edge-based service carried over IP, and thus > is not "in" the core any more than any other protocol running > over IP that the core happens to be carrying. P routers should > not have a clue that L2TP exists. When I say leave L2TP in the core, I mean continue to use L2TP on the PE core facing interfaces to provide PE-PE tunnels across the core, as opposed to using MPLS PE-PE tunnels across the core. > > This isn't analogous to the IP/GRE case as these don't have > > their own muxing capabilities and therefore MPLS is used for > > this purpose. > > Just to be clear, GRE has a muxing capability via its 32-bit Key. And IP has a muxing capability via its UDP/TCP port numbers. I was referring here to the use of muxing fields as defined by the IETF PWE3 WG for VPN muxing. These are currently MPLS or L2TP session IDs (and maybe IP UDP port numbers, I've lost track of whether on not this has been accepted by PWE3 as a valid approach). The point being that PWE/L2VPN L2TP based solutions exist today that do not require MPLS for muxing, whereas I am not aware of any PWE3/L2VPN IP/GRE solutions that do not require MPLS labels for muxing. Regards, Richard _______________________________________________ mpls mailing list mpls@lists.ietf.org https://www1.ietf.org/mailman/listinfo/mpls |
|