Service Provider & MPLS
L3VPN
The flagship MPLS service: BGP/MPLS IP VPNs (RFC 4364), where one shared backbone carries thousands of customers' overlapping address spaces in isolated routing tables. Two chapters: the architecture — VRFs, route distinguishers, route targets — and the end-to-end control and forwarding path.
- read in order, or dip in
- 2 chapters
- updated 26 Oct 2026
- Free — no signup
read in order, or dip in
updated 26 Oct 2026
Contents
- 01MPLS L3VPN Architecture: VRFs, RDs and Route TargetsAn MPLS L3VPN isolates each customer in a VRF — a private routing table on the PE router — while the shared core stays customer-blind. Route distinguishers make overlapping customer prefixes globally unique inside BGP (RD + prefix = VPNv4 route), and route targets — extended communities exported and imported per VRF — control which sites learn which routes, making topologies from any-to-any to hub-and-spoke a matter of import/export policy.
- 02MP-BGP and the L3VPN Forwarding PathL3VPN routes travel in the VPNv4 address family of MP-BGP between PE loopbacks: each route carries its RD-prefixed prefix, route targets and a VPN label allocated by the egress PE. Forwarding stacks two labels — an LDP transport label to reach the egress PE and the BGP VPN label beneath it identifying the VRF — so a packet crosses a customer-blind core and lands in exactly the right customer table.
Lab guides
9 guides · 4 downloadable
- MPLS - Layer3 VPN with OSPF Protocol between CE-PEPDF
- MPLS - Layer3 VPN with RIP protocol between CE-PEOpen
- Complex MPLS VPN with BGP Protocol between CE-PEPDF
- MPLS - Complex MPLS VPN with OSPFPDF
- Complex MPLS VPN with OSPF on CE-PE RoutersPDF
- MPLS - Multi-VRF or VRF-LiteOpen
- MPLS - Sham-Link Support on Multi-VRFOpen
- NCS Lab 5: Configuring Intra-AS MPLS L3VPN using Static Routing as PE-CE ProtocolOpen
- NSO Lab 6: L3VPN (Layer 3 Virtual Private Networks)Open
Spotted something out of date? Tell us — this library is maintained by the faculty.