The topology is composed of 8 routers and they are actually Cisco 3640 routers. It includes 3 simulated ISPs (AS 30, 50 and 100) and 2 simulated customer (AS 200 and 192) that are multi-homed into two different ISPs.The 2 customer routers are multi-homed so that they can have 2 different paths to get to each other, one path is passing through the AS 30 and 50 and the other one is through AS 100. Leaving the BGP configurations by default the 2 customer routers will choose the path through AS 100, and that is because of the BGP AS-Path attribute. A BGP router will choose a path to a certain destination with a lower AS path/hop-count. Therefore, R0 will pass through AS 100 to get to R5 since it is only one hop away from itself and this is also true for R5 going to R0.
I chose this kind of setup so that we can have a chance to manipulate the BGP attributes and for us to be able to control the traffic. Like for example the Local Preference attribute, we can tell R0 to use the path going to AS 30 and 50 instead of AS 100 by setting the local preference of the routes that came from AS 30 to have a value greater than those that came from AS 100.
For the ISPs, the routers inside their respective AS are in full mesh IBGP configuration, which is actually the requirement and is what ISPs really practice in their backbone. In AS 30, R1 and R2 are peering to each other via their Loopback IP address but in order for these two routers to have a successful neighbor relationship they must have Layer 3 reachability to each other's Loopbacks. In order to accomplish this we need a routing protocol that will tell the routers how to reach other's Loopbacks, in this case we will used OSPF routing protocol to do the job and this is also how we will setup the other 2 simulated ISPs. Also, we will redistribute only the Loopback interfaces of the routers for us to be able to verify the connectivity and advertisements in between ISPs but we can still redistribute all the connected interfaces if you want to, anyway this is a Lab we can do whatever you want.
Peering in between different ISPs and in between customers and ISPs, of course, are implemented using EBGP, R2 and R3 are using EBGP to peer with each other. But how do we really know or determine if a router is using IBGP and/or EBGP?
Answer:
Configuration Example:
router bgp 30
neighbor [neighbor ip] remote-as 30
neighbor [neighbor ip] remote-as 50
The line "neighbor [neighbor ip] remote-as 30" indicates IBGP peering, meaning we peered with a router that is in the same AS number. Now, the line "neighbor [neighbor ip] remote-as 50" indicates EBGP peering, meaning we peered with a router that is outside or in different AS number.
So, basically that is all with the topology, in my next blog I will post all the router configurations so that you can really implement the network topology in our PCs at home.
