Scenario:
As the senior network engineer for a large international company you are always trying to optimize your network and get the most out of your time. The network has a lot of IBGP routers and you got sick of configuring all those neighbor peerings. To fix this you configured the network with some route reflectors…this has saved you a lot of work but you want to take it one step further by clustering the network.
Goal:
- All IP addresses have been preconfigured for you. You don’t see them in the topology picture or it would look like a picasso painting.
- Each router has a loopback0 interface.
- All routers belong to AS 1.
- Configure router Tilburg as the route-reflector for router Amsterdam and Utrecht. Use cluster-ID 1.1.1.1.
- Configure router NewYork as the route-reflector for router LosAngeles and Boston. Use cluster-ID 4.4.4.4.
- Configure router Hamburg as the route-reflector for router Dusseldorf and Berlin. Use cluster-ID 7.7.7.7
- Configure router Tilburg,Newyork and Hamburg with the correct IBGP neighbor peerings.
- Advertise all loopback0 interfaces in BGP.
- Ensure you have full reachability within AS 1. All loopback0 interfaces should be reachable.
IOS:
c3640-jk9o3s-mz.124-16.bin
Topology:
Video Solution:
Configuration Files
You need to register to download the GNS3 topology file. (Registration is free!)Once you are logged in you will find the configuration files right here.
Do you want your CCNA or CCNP Certificate?
The How to Master series helps you to understand complex topics like spanning-tree, VLANs, trunks, OSPF, EIGRP, BGP and more.
Written by René Molenaar - CCIE #41726
Are the goals below the topology an error? It looks like they belong to another lab 🙂
Hi Mattias,
The goals are ok. what do you think is confusing? 🙂
Rene
Do you have solution for this lab?
I don’t have the final configuration for this one yet. In a few weeks i’m back to recording everything. If you are stuck on some part please leave a comment here or at the forum, I’ll answer it ok?
When I try and load this topology, it says that the device Boston is non-existent. Is anyone else having the same problem? I’ve tried removing and re-adding multiple times and the same results everytime.
I just opened it on my PC and it’s working here. Is anyone else having issues with this one? I can vaguely recall that I edited this topology after creating it, maybe there’s something wrong with it.
If others have the same issue I’ll recreate it.
Hello Rene,
there is mistake in configuration for LosAngeles FastEthernet1/0:
>>ip address 192.168.46.4 255.255.255.0
it should have ip address from 192.168.56.0/24 network
I’ll fix it 🙂
Hi rene,
why? your are using clusters-id
when you can complete the lab without clusters,
what is the main purpose of using Cluster ?
You could do it without clusters but his lab is created to demonstrate how route reflectors and cluster ID works. Here’s a good article from IPexpert that explains it a bit more:
http://blog.ipexpert.com/2012/02/20/understanding-bgp-originator-id-and-cluster-id/
The cluster-id that you have in this scenario is incorrect. Each router should have it’s own cluster-id #. If you shutdown the link between L.A and New York, the router in L.A. wouldn’t fail over to Boston. It’s only when the cluster-id is unique, the fail over situation will work. This is documented on Cisco website and on the below link
http://rekrowten.wordpress.com/2013/05/24/bgp-route-reflector-and-why-is-cluster-id-obsolete/
good lab.
This Lab cannot be completed if only BGP is used in this lab: I have tested it two ways:
1. Configured BGP as required and advertised the interconnected networks into all three route reflectors into BGP. After that you will learn the routes to all the loopbacks and they are marked as valid and best but they will not be ping because the next hops are unreachable. Their is no underline transport available for the next hops. If your reach-ability means that the routes should be in the routing table then it is fine but for me reach-ability mean they should be ping from everywhere…
2. I removed the transit links from bgp and only advertised ONLY the transit links of all routers into OSPF. The loopbacks were still advertised into BGP. Only this way i was able to ping all the loopbacks sourcing from loopback
Hi Rene
In my case, Dusseldorf was not receiving prefixes outside of cluster 7.7.7.7. While Berlin was receiving all prefixes. As I changed cluster id to 11.11.11.11 (previously 7.7.7.7), Dusseldorf started to receive prefixes.
Don’t know if is this a bug or someone else experienced the same issue?
Thanks.
Another question, why does Tilberg, NewYork and Hmaburg acts as route-reflector-clients for each other?
Thanks.