Earlier in this guide, we learned that, by default, a BGP speaker will not any advertise routes received from an iBGP peer to another iBGP peer. While this default behavior is designed to prevent loops in iBGP implementations, it also imposes scalability issues. If, for example, a network has four routers running BGP, then all routers must be peered to each other in order for all routers to learn about the iBGP prefixes. This means that each BGP speaker must have three neighbor statements – one for every other router, resulting in twelve neighbor relationships for BGP. As the number of . . .
The content below this message is for members only.