BGP Route Reflector

Route reflectors (RR) are one method to get rid of the full-mesh of IBGP peers in your network. The other method is BGP confederations.

The route reflector allows all IBGP speakers within your autonomous network to learn about the available routes without introducing loops. Let me show you an example picture:

IBGP 6 routers full mesh

Above we have a network with 6 IBGP routers. Using the full mesh formula we can calculate the number of IBGP peerings:


So that will be:

6(6-1=5) / 2 = 15 IBGP peerings.

When we use a route reflector our network could look like this:

IBGP Route Reflector Example

We still have 6 routers but each router only has an IBGP peering with the route reflector on top. When one of those IBGP routes advertises a route to the route reflector, it will be “reflected” to all other IBGP routers:

IBGP Route Reflector Route Update

This simplifies our IBGP configuration a lot but there’s also a downside. What if the route reflector crashes? It’s a single point of failure when it comes to IBGP peerings. Of course there’s a solution to this, we can have multiple route reflectors in our network. I’ll give you some examples later.

The route reflector can have three type of peerings:

  • EBGP neighbor
  • IBGP client neighbor
  • IBGP non-client neighbor

When you configure a route reflector you have to tell the router whether the other IBGP router is a client or non-client. A client is an IBGP router that the route reflector will “reflect” routes to, the non-client is just a regular IBGP neighbor.

When a route reflector forwards a route, there are a couple of rules:

  1. A route learned from an EBGP neighbor can be forwarded to another EBGP neighbor, a client and non-client.
  2. A route learned from a client can be forwarded to another EBGP neighbor, client and non-client.
  3. A route learned from a non-client can be forwarded to another EBGP neighbor and client, but not to a non-client.

The third rule makes sense, this is our normal IBGP split horizon behavior.

Now you have an idea what the route reflector is about, let’s take a look at some configurations.


We’ll use a simple example, 3 IBGP routers with a single route reflector:

AS 123 IBGP Route Reflector

In this example we have 3 IBGP routers. With normal IBGP rules, when R2 receives a route from R1 it will not be forwarded to R3 (IBGP split horizon). We will configure R2 as the route reflector to get around this. Let’s configure R1 and R3 first:

R1(config)#router bgp 123
R1(config-router)#neighbor remote-as 123
R3(config)#router bgp 123
R3(config-router)#neighbor remote-as 123

The configuration of R1 and R3 is exactly the same as a normal IBGP peering. Only the configuration on the route reflector is special:

R2(config)#router bgp 123
R2(config-router)#neighbor remote-as 123
R2(config-router)#neighbor route-reflector-client

R2(config-router)#neighbor remote-as 123
R2(config-router)#neighbor route-reflector-client

Here’s the magic…when we configure the route reflector we have to specify its clients. In this case, R1 and R3. In my topology I have added a loopback interface on R1, let’s advertise that in BGP to see what it looks like on R2 and R3:

R1(config)#router bgp 123
R1(config-router)#network mask

That’s all we have to configure. Let’s use some show commands to verify our work.


First we’ll look at R2, see if it learned anything:

R2#show ip bgp
BGP routing table entry for, version 2
Paths: (1 available, best #1, table Default-IP-Routing-Table)
Flag: 0x820
  Advertised to update-groups:
  Local, (Received from a RR-client) from (
      Origin IGP, metric 0, localpref 100, valid, internal, best

R2 shows us that this route was received from a route reflector client. Did it advertise anything to R3? Let’s find out:

R2#show ip bgp neighbors advertised-routes
BGP table version is 2, local router ID is
Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,
              r RIB-failure, S Stale
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network          Next Hop            Metric LocPrf Weight Path
*>i1.1.1.1/32             0    100      0 i

So what do we see here? Let me explain…

We're Sorry, Full Content Access is for Members Only...

If you like to keep on reading, Become a Member Now! Here is why:

  • Learn any CCNA, CCNP and CCIE R&S Topic. Explained As Simple As Possible.
  • Try for Just $1. The Best Dollar You've Ever Spent on Your Cisco Career!
  • Full Access to our 662 Lessons. More Lessons Added Every Week!
  • Content created by Rene Molenaar (CCIE #41726)

501 Sign Ups in the last 30 days

100% Satisfaction Guaranteed!
You may cancel your monthly membership at any time.
No Questions Asked!

Forum Replies

  1. Hi Rene,
    Ur explanation was good…plz also describes about BGP Route Reflectors and Confederations.

  2. Hi Hamood,

    Good question, there’s a good explanation for this:

    1. iBGP requires a full mesh of peerings because of iBGP split horizon. This is why we use loopback interfaces instead of physical interfaces for the peering. Physical interfaces can go down, loopbacks can’t (unless you shut them). In this example I could have used physical interfaces since there is only one link between R2-R3 and R3-R4, if we had a link between R2-R4 then it would have been a must.

    2. When R4 (or R2) advertises its network on the loopback interface to R3 then R3 will store it in its

    ... Continue reading in our forum

  3. Hello Minh,

    It sounds like you have a good understanding of these concepts :slight_smile:

    ISPs / service providers often use MPLS in their core networks yes. One of the advantages is that you don’t have to run iBGP on each and every core router. I have a lesson where I explain this:

    ... Continue reading in our forum

  4. actually it is helpful. I have ran into that in the work place. I just was so focused I was not seeing it. I have actually ran into that in the work environment. where we had a bgp prefix of a customer and we needed to test to see if they could get outside the ISP network. My initial pings did not work because of something similar here where the IP being used by default by the ping was not the correct one and we had to use the source command. So that is vey similar to this except in those cases it was actual public IPs and not loopbacks. However, now

    ... Continue reading in our forum

150 more replies! Ask a question or join the discussion by visiting our Community Forum