The key to understanding the general idea of how MPLS VPNs work is to focus on the control plane distinctions between PE routers and P routers. Both P and PE routers run LDP and an IGP to support unicast IP routing. However, the IGP advertises routes only for subnets inside the MPLS network, with no customer routes included. As a result, the P and PE routers can together label switch packets from the ingress PE to the egress PE.

PEs have several other duties as well, all geared toward the issue of learning customer routes and keeping track of which routes belong to which customers. PEs exchange routes with the connected CE routers from various customers, using either EBGP, RIP-2, OSPF, or EIGRP, noting which routes are learned from which customers. To keep track of the possibly overlapping prefixes, PE routers do not put the routes in the normal IP routing table—instead, PEs store those routes in separate per-customer routing tables, called VRFs. Then the PEs use IBGP to exchange these customer routes with other PEs—never advertising the routes to the P routers.

Both ASBR1 and ASBR2 advertise defaults into the network, expecting to have the capability to route to the Internet through BGP-learned routes. In this case, ASBR2 is already up, fully converged. However, if ASBR1 reloads, when it comes back up, OSPF is likely to converge faster than BGP. As a result, ASBR1 will advertise its default route, and OSPF routers may send packets to ASBR1, but ASBR1 will end up discarding the packets until BGP converges.

Using the stub router feature on the ASBRs solves the problem by making them advertise infinite metric routes (cost 16,777,215) for any transit routes—either for a configured time period or until BGP convergence is complete. To do so, under router ospf, the ASBRs would use either the maxmetric router-lsa on-startup announce-time command or the max-metric router-lsa on-startup wait-for-bgp command. With the first version, the actual time period (in seconds) can be set. With the second, OSPF waits until BGP signals that convergence is complete or until 10 minutes pass, whichever comes first.

Frame-relay switching is simple to configure. In this example we have 2 serial ports and we want to be able to switch dlci’s between them. Serial port 1/0 connects to customerA using DLCI 100 and Serial 1/1 connects to customerB using DLCI 200. Both customers need reachability to eachother.

Simply enable frame-relay switching on your router.

# frame-relay switching

Then enable the ports for frame-relay

# int ser1/0

# encapsulation frame-relay

# int ser1/1

# encapsulation frame-relay

We need to specify these ports to be DCE even though they may be DTE interfaces.

# int ser1/0

# frame-relay intf-type dce

# int ser1/1

# frame-relay intf-type dce

Now we have two options. We can use the old school route method or the modern connect method.

Old School: under interface

# int ser1/0

# frame-relay route 100 interface serial1/1 200

# int ser1/1

# frame-relay route 200 interface serial1/0 100

Modern: Global config

# connect custA_custB serial1/0 100 serial1/1 200

As you can see the modern “connect” method is much simpler and uses less config. And don’t forget to put in the clock-rate if they are indeed DCE interfaces.

Cheers.