Multiple BGP route reflectors question

[Hassaan]

Hi all,

I'm looking for advice on how to configure a network with more than one route reflectors.

My specific scenario is that I wish to redistribute an ebgp route into our AS and within this AS I wish to create ibgp route reflectors to "reflect" that route to all routers withing my AS. 

I've set it up so that the border routers (in my AS) are configured as route reflector clients and then an RR situated one hop down reflects that to other RR clients across the network. 

The redistribution and reflection part works fine, now I want to set up two RRs for failover, but an then an issue arises. The issue is that the ebgp route does not get advertised to the other RR.

The two RRs have a direct physical connection, but I also created an ibgp peering between them.

I tried to troubleshoot this by:

- giving the two RRs the same cluster ID

- giving the two RRs different cluster IDs

- I tried to make the two RRs clients of each other

None of the above worked.

Any suggestions?

[Flavio Miranda]

 bgp client-to-client reflection all

[Georg Pauwen]

Hello,

post a diagram of your topology as well as the full configurations (sh run) of all routers involved...

[MHM Cisco World]

The eBGP is connect to only one RR not to both? If Yes then you need to add One router that connect to eBGP one side and other side connect to both RR 
this additional Router will be client for both RR and hence will forward the prefix to both since it client for Both. 
RR receive the prefix and re advertise it to other Lower client.

[Hassaan]

Below is my topology and I've attached the running config of the two RRs. Let me know if you would like me to add the client router configs.

@Flavio Miranda I applied that command on the two RR routers and it didn't seem to have any effect. It didn't even appear in the configuration, is that expected?

@MHM Cisco World I see what you are saying and that seems like one way that could fix the issue. But, in my topology eBGP is coming in to both RRs and I want to reflect the IP of SP-1 and SP-2 to all routers within AS 10 for full resilience, e.g. if client-2 were to have an outage all routers will still be able to access the Internet via client-1. Also, in your diagram wouldn't that non-client router reflecting between the two RRs be a single point of failure? What I did do was create an iBGP peering between client-1 and both RR-1 and RR-2 and likewise for client-2 (this is not shown in the attached running config). That did work and advertised the 194.82.96.1 and 194.82.98.1 to both RRs. Is this the only way to do it? Or is there a simpler way?

Note that I haven't added client-3 and client-6 as peers yet.

[Georg Pauwen]

Hello,

this looks like a GNS3 project ? If it is, post the entire project, and indicate which IOS images you are using, and in which local directory the images are stored.

[paul driver]

Hello

i suggest to use different cluster id on each RR this will negate reflected routes  from either RR from being ignored

Lasty both RR are just ibgp peers netween each other and not RR clients

[Hassaan]

Thanks @paul driver ,

I did try a both your suggestions at the beginning and it didn't work. Just tried it again to be sure and still didn't work.

If the only way to get around this problem is to use @MHM Cisco World solution, enhanced with a backup non-client (or client) router then that will have to suffice, but not ideal. 

[MHM Cisco World]

I will do lab tonight and send you what is best solution if there is other than ADD router between ISP and RR's.

[MHM Cisco World]

[paul driver]

Hello @MHM Cisco World 

can you elaborate on this please which is/are the RRs?

Do you have physical resilience  for the RRs?

[MHM Cisco World]

@paul driver 

there are two RR in my topology "please refer to photo I attach"
and Yes there is Physical direction connection between the two RR.

[paul driver]

Hello

Okay i can see that now,  I see with the design you have no resiliency for clients 1/2, 4/5 and the single rtr that does have resiliency its not clear what client rtr from the OP it represents?
Lastly you still have non-clients which i assume will be 3/6 (not shown)

[paul driver]

Hello

---

@Hassaan wrote:

 did try a both your suggestions at the beginning and it didn't work. Just tried it again to be sure and still didn't work.

---

My suggestion was based on your OP but now you have posted a topology diagram, then I can see why it wasnt a viable suggestion.

What you can do is have four RR within two clusters, this way you would have full resiliency of the RRs, also all RRCs would have dual connection to each of their respective RRs within their own cluster, then any ebgp routes should be successfully reflected to all rtrs you shouldn't require any additional router either, just re patching would be applicable

[ahmadaljarah]

optimal Design is not to use the ASBR as RRS and to transport  VRF Prefixes you need VPNV4 address-family which is not there at your config, or you can advertise default route at ipv4 BGP and  Leak the Routes to VRF  at RRS