3

I cant understand one thing, in this scenario:

enter image description here

This is an schema of a layer 3 vpn using mpls. I dont understand why devices conncted to VPNa and VPNb can have the same addresses, I mean, how can PE's differentiate it? I know a little and I know that they use a routing table called VRF, but I dont undertand it, If I am in VPNa with an IP of 172.0.2.5 and there is another device on VPNb with that address (172.0.2.5), how can PE's distinguish that traffic and differentiate it? Thanks I google it a lot of times but I could not find and answer.

4

When traffic arrives into a PE router from a customer the interface that it arrives on is associated with a VRF. The router will only lookup the destination in the VRF routing table for the associated VRF. If the interface is associated with VRF A, the router will only be able to see the 172.0.2.5 that exists within VRF A. If the interface is associated with VRF B, the router will only be able to see the 172.0.2.5 that exists within VRF B. The router will either forward the traffic locally out of another interface associated with the same VRF, or if the destination is on another PE router the traffic is encapsulated in MPLS and forwarded to the egress PE router.

If you are familiar with VLANs, VRFs are very similar, but at L3 instead of L2. When a frame arrives on switch, the port is associated with a VLAN and the switch looks up the destination MAC address in the MAC address table for that VLAN. The same concept exists for VRFs at layer 3. When a packet arrives on a router, the interface is associated with a VRF and the switch looks up the destination IP address in the routing table for that VRF and then forwards out of that interface (which is also in the same VRF).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.