Lets try tp explain this with an example
NetworkA----R1---R2---R3--networkB
Say R1 and R3 runs IBGP. R1, R2, R3 run OSPF between them.
Now when the update is sent to R3 from R1, the next hop is indicated as R1. So when a packet destined for NetA arrives at R3, it does a route lookup abd finds out that the network is learned by IBGP and the next hop is R1. Now it will do a recursive lookup to find out the route to the next hop R1 which will be learnt by OSPF and thus sends the packet to R2. when the packet arrives at R2, the destination does not match any prefixes on its OSPF table and hence it gets discarded
So if you decide to run IBGP instead of OSPF as your IGP, then you require a ful mesh IBGP. If you just configure the network such that R1 is the IBGP peer to R2 and R2 is IBGP peer to R3, it does not work.
This is because When routes are advertised via IBGP, they are by definition advertised within the same AS. As a result, the AS_PATH does not change. As a result, the IBGP routes do not have the loop protection that EBGP routes have. To protect against loops, BGP does not advertise routes that have been learned from an IBGP peer to another IBGP peer. Thus R2 does not advertise the networks learnt via R1 to R3.
To overcome this you create a Full IBGP mesh between all the 3 routers i.e R1 speaks BGP with both R2 as well as R3
The best option though is to let the IGP use the deafult route and forward it to the nearest edsge router where BGP can take the decision for routing on the internet
HTH
Narayan
Please rate all posts
