IBGP peering with Two hope far from RR

[Deepak Kumar]

Hi,

Yesterday, I was working in my lab and found some strange issues with BGP. My LAB setup is like:

1. R1 is adverting the default route in BGP which is configured toward the R2. (I know this is strange but it was my lab).

2. All Loopbacks are advertised in the BGP.

3. Point to Point subnets are not advertised in the BGP and there is no IGP routing protocol.

4. R3 is RR and R8 and R3 are not connected directly so I am trying to configure directly peering between R3 ad R8 to use RR feature.

R8 Routing table:

B* 0.0.0.0/0 [200/0] via 192.168.4.1, 00:11:44
1.0.0.0/32 is subnetted, 1 subnets
B 1.1.1.1 [200/0] via 192.168.4.1, 00:11:44
2.0.0.0/32 is subnetted, 1 subnets
B 2.2.2.2 [200/0] via 192.168.4.1, 00:11:44
8.0.0.0/32 is subnetted, 1 subnets
C 8.8.8.8 is directly connected, Loopback0
192.168.4.0/24 is variably subnetted, 2 subnets, 2 masks
C 192.168.4.0/24 is directly connected, FastEthernet0/0
L 192.168.4.2/32 is directly connected, FastEthernet0/0
B 192.168.99.0/24 [200/0] via 192.168.4.1, 00:11:44

BGP Configuration on the R8:

router bgp 4500
bgp router-id 8.8.8.8
bgp log-neighbor-changes
network 8.8.8.8 mask 255.255.255.255
neighbor 192.168.2.2 remote-as 4500
neighbor 192.168.4.1 remote-as 4500

RR (R3) configuration:

router bgp 4500
bgp router-id 3.3.3.3
bgp log-neighbor-changes
network 3.3.3.3 mask 255.255.255.255
neighbor 192.168.2.1 remote-as 4500
neighbor 192.168.2.1 route-reflector-client
neighbor 192.168.2.1 next-hop-self all
neighbor 192.168.3.2 remote-as 4500
neighbor 192.168.3.2 route-reflector-client
neighbor 192.168.3.2 next-hop-self all
neighbor 192.168.4.2 remote-as 4500
neighbor 192.168.4.2 route-reflector-client
neighbor 192.168.4.2 next-hop-self all

Ping From R3 to R8:

R3#ping 192.168.4.2 repeat 2
Type escape sequence to abort.
Sending 2, 100-byte ICMP Echos to 192.168.4.2, timeout is 2 seconds:
!!
Success rate is 100 percent (2/2), round-trip min/avg/max = 8/12/16 ms
R3#

Ping From R8 to R3:

R8#ping 192.168.2.2 repeat 2
Type escape sequence to abort.
Sending 2, 100-byte ICMP Echos to 192.168.2.2, timeout is 2 seconds:
!!
Success rate is 100 percent (2/2), round-trip min/avg/max = 12/30/48 ms
R8#

Error message on both routers:

*Oct 1 22:17:31.401: BGP: 192.168.4.2 Active open failed - no route to peer, open active delayed 9216ms (35000ms max, 60% j itter)
*Oct 1 22:17:40.617: BGP: 192.168.4.2 Active open failed - no route to peer, open active delayed 7168ms (35000ms max, 60% j itter)
*Oct 1 22:17:47.785: BGP: 192.168.4.2 Active open failed - no route to peer, open active delayed 7168ms (35000ms max, 60% j itter)

Why is it happening: As I know why is this happening because IBGP having simple rules as

R8 & R3 needs a static route to the remote peering link, or R2 must advertise the P2P network into BGP. Hence I do not enter the static route because I have a default route.

What is my Question:

As mentioned that default route is there and I can reach the remote peering IP address so the static route is not looking mandatory (as based on IP connectivity rules). If BGP having rule that peering IP/Subnet must be routing table then here is something missing as:

If I will add a static route on R8:

IP route 192.168.2.0 255.255.255.0 192.168.4.1

then BGP neighborship is establishing without any issue but still, R3 is not having 192.168.4.0/24 subnet in it's routing table. As per rule, it will also ask for a static route.

Looking some technical guide and deep drive on this case.

[paul driver · Accepted solution]

Hello
What your seeing suggests its due to the default routes, The ip connectivity is there but because of these defaults BGP doesn't trust it, Its going to perform a route lookup for its peer address and if it only finds a default route in its route table then the bgp peer wont begin to establish.

If you think about as the default just points egress it makes sense because otherwise that bgp peer could be anywhere.

[Deepak Kumar · Accepted solution]

Hi @paul driver ,

Thanks for the answer. But my question is why remote peer (R3) which is RR is having an only default route in the routing table but still, it is getting neighborship up. Why only one side of the peer is caring about the default route?

For more explain:

1. R3: Having default Route and R8: having default route= No neighborship getting up

2. R3: Having a static route and R8: having to default route= Neighorship UP

3. R3: Having a default route and R8: Having static Route= Neighorship UP

Point 2 and Point 3 are points where I am getting disappointment. Why one side BGP Peer is caring about the Static or default route?

[paul driver]

Hello

I believe that those static routes are providing the sanity check via the route lookup that bgp requires so those rtrs can initiate/establish the connection

[Deepak Kumar]

Hi,
Ok Thanks