Using route reflectors
We will now enable (on a very limited scale) route reflection in BGP. To do this, we will be reusing the pod policy we created earlier (PoD-Policy
), back in Chapter 2.
This is the same method you would use if you wanted to implement multiprotocol BGP. Strictly speaking, however, route reflection is not the same as multiprotocol BGP (MP-BGP). Route reflection reflects routes from one route reflector client to another, through a "server," bypassing the need for a full mesh between BGP speakers. MP-BGP is used to carry both IPv4 and IPv6 traffic (as well as MPLS VPN traffic), using "address families." The two can be implemented together; you can have route reflection within MP-BGP, but to say that configuring BGP route reflectors also implements MP-BGP is not quite correct. We are looking at this from an ACI standpoint, though, so let's refer to this as an "alternative fact" and move on to configuring MP-BGP.
Within ACI, MP-BGP serves to distribute the routes received from...