The provider’s MPLS cloud has three routers namely – R1 (P-router), R2 (PE-R2) and R3 (PE-R3). These routers formed OSPF adjacency with one another. R2 and R3 are iBGP neighbors peering with each other’s loopback address.
The TTL propagation within the MPLS cloud was suppressed with
no mpls ip propagate-ttl
command. This is to “hide” the number of mpls routers that exist within the provider’s MPLS core.Reason to use ospf sham link
It is possible that customer’s network has an OSPF backdoor link to each other despite subscribing MPLS service which links customer’s edge routers.
The OSPF link through the MPLS cloud would be an inter-area link despite both site-a and site-b links are in OSPF area 0, this poses a problem if customer wants traffic to traverse from site-a to site-b or vice versa through the MPLS core. OSPF will prefer the intra-area route, in this case is the backdoor link which resides in the same OSPF area, to reach the destination.
To solve this problem, OSPF sham link is used.