Hi there:
While studying BGP, a doubt came up , on weather in production environment the ISP would advertise its internal(IBGP) networks to an EBGP neighbor. The fact that only the edge router(R3) can ping advertised network form R4, due to the fact that it knows the next hop of the advertised routes and R4 knows how to get back to R3. R2 and R1 would not be able to ping, because R4 does not know how to get back to them. Is this seen in production networks?
By the way, love the teaching style and happy learning.
The advantage of BGP is the fact that it advertises routes based on AS path, and not on next hop. In a sense, BGP routes from AS to AS rather than from router to router. This makes it more scalable to be able to deal with the vastness of the Internet.
When an eBGP router, such as R3, exchanges routes with the customer edge router R4, it will only advertise routes to other ASes on the Internet. Typically, it is not necessary for R3 to inform R4 of any internal prefixes, such as those of R1 and R2. The only reason R3 would inform R4 of any destinations internal to AS 100 in your diagram, is if the customer needs to reach some resource within the ISP’s AS, such as a web server, a DNS server, or any other service.
Otherwise, I agree with you, the ISP would not typically share any prefixes internal to its own AS with the customer premises devices.
