Consider the following scenario: An AS with multiple (lots) internal routers exchanging information via EIGRP. Out of these routers, two are border, with EBGP feeds to upstream providers. One of these routers is used as a default entry/exit point, whilst the other is used to lab purposes. A third router, somewhere in the depth of the network, requires a BGP routing table.
Neighbors (ibgp) between this third router and the "lab" border router have been established, with a lovely set of Internet routes reciding on the third router. The next hop of the ibgp routes are pointing towards the loopback of the "lab" border router. Everything seems absolutley fine. All loopbacks are reachable via the IGP, and from any BGP router.
Now, When I trace from the third router (from an interface address which is been announced via the "lab" border router to upstream) it decides to follow the candidate default route located on next-igp-hop (a pure igp router), exiting via the wrong border router, not the one specified by the BGP NEXT HOP.
I've checked the administrative distances, and they seem to be fine (perhaps not).
I'm confused, why could this be happenening? Any feedback would be greatly appreciated.
I assume, the 3rd IBGP router is peering with the lab router's loopback interface through multiple IGP hops in between them. Your gateway on the 3rd IBGP router might be pointing to the lab router but the non-bgp (eigrp) routers on the path might be using a default route to point to the primary EBGP router and hence, the routing doesn't the work the way you intend it to be. That's the reason one of the BGP requirement is to have full mesh or workaround that using confederation/route reflector configuration. Another option is to redistribute BGP into EIGRP and that mightn't be practical if you are learning full/partial internet routing table from your ISPs.
RRs are ideal for a small BGP network. However, if your's is an enterprise network, then you would have to go for confederation (sub-ASs) within your AS.
If you are unable to redistribute the BGP routes into EIGRP then you would have to create a full mesh using one of the two techniques mentioned above for all the routers to choose the optimal path out to the Internet.
But, doesn't confederation make the BGP peering ideal for large scale networks? Wouldn't breaking a AS into many sub-ASs make it easy to manage. Ofcourse, the admin would have to be comfortable with the rules of BGP in a confed environment. Sure, you could have RRs within the sub-ASs. A flat AS with no sub-ASs and just RRs doesn't sound like a great idea to me.
I have seen networks with close to a thousand nodes running RRs with no issue whatsoever. The argumentation about confederation being more scalable than RR has been put aside in favor of simplicity.
You can scale with RR too using multiple level of RRs if need be.
I doubt that an entreprise network would ever need to consider confederation since I have never ever seen an entreprise network running iBGP on all routers in the AS. That sounds to me like it would constitute a complex network to operate, which at best would potentially be suitable for an SP with experienced operation folks but rather overwhelming for an entreprise customer.
Harold Ritter Sr. Technical Leader CCIE 4168 (R&S, SP) email@example.com México móvil: +52 1 55 8312 4915 Cisco México Paseo de la Reforma 222 Piso 19 Cuauhtémoc, Juárez Ciudad de México, 06600 México
Hi everyone, I would like to thank you in advance for any help you can provide a newcomer like myself!
Im studying the 100-105 book by Odom and am currently on the topic of Port security. I purchased a used 2960 and I'm trying to follow a...
While deploying a number of 18xx/2802/3802 model access points (APs), which run AP-COS as their operating platform. It can be observed on some occasions that while many of their access points were able to join the fabric WLC withou...
I am going to design and build an LAN network under a tunnel underground with long distance between the switches.
I will have 2 Catalyst switches and 8 Industrial IE3000, and they will be connected with fiber.
For now I am planning on use Layer-2 s...