cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1539
Views
0
Helpful
12
Replies

Dual homed EBGP issue

Yangjp715
Level 2
Level 2

Hi all,

 

I build a test environment with IBGP and EBGP. Please see the following diagram. There are two links on R5, one is connected to R1, another is connected to R2. Both links are active with the same cost. All IP addresses of loopbacks and interconnect are advertised under OSPF and BGP. R2 and R4 are BGP route reflector, R1 and R3 are RR client for both RR. The issue is that 2.2.2.2 and 4.4.4.4 are not reachable from R5 if the both links are connected. R2 and R4 are pingable from R5 if the link which is connected to R3 is disconnected. R2 and R4 are not pingable from R5 if the link which is connected to R1 is disconnected. 

 

Is there anyone who can fix this issue?

 

Thanks in advance!

Eric

 

2017-11-27 13_35_44-Test environment.png

12 Replies 12

Hello

You physical topology doesn't look correct.

R1/R3 are redistribution rtrs between ospf and bgp but they have only one physical connection for both bgp and ospf doesn't make sense .

can you post the config for R1/R3 please

res
Paul

 

 


Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.

Kind Regards
Paul

Please see the attached.

R3

Hello,

 

I labbed this in VIRL. The problem is with R4:

 

GigabitEthernet0/0.10 (VLAN 10) has a different address than the same VLAN 10 interface on R2.

Change the entire subinterface on R4 to GigabitEthernet0/0.20 (VLAN 20). Also.
make sure that all three VLANs exist on bith switches. And change GigabitEthernet0/1
on SW2 to 'switchport access vlan 20'.

 

If you don't get it to work, I'll send over the working configs...

OSPF p2p:
G0/1 on R1 and GigabitEthernet0/0.10 on R2.
GigabitEthernet0/0.30 on R2 and GigabitEthernet0/0.30 on R4
G0/1 on R3 and GigabitEthernet0/0.10 on R4.
VLAN 10 is enabled on SW1 and SW2.

Hello,

 

here are the working configs. Check in how far yours are different...

And the switches:

 

 

It is almost the same configurations. The only difference is RR postion. It is interesting, when i configured R1 and R3 as RR, i can ping R2 and R4 from R5, but R5 is not reachable from R2 and R4.

Now R5 is not reachable from R2 and R4. but can be reached from R1 and R3.

lespejel
Level 3
Level 3

You should consider to rebuild your L2 desing, from R1 to R1 you use Vlan 10, but also from R2 to R4, and R3 to R4.

Have you set root devices for STP? check port status and see if any port is blocked.

 

My recommendation is to set a vlan from R1 to R2, one from R2 to R4, and one for R3 to R4, even when you are using different IP addresses, you need to check if L2 ports state is active or forwarding.

 

And I don't see the need to have your vlan 10 extended over 2 devices, remove it from the trunk between switches, that might help.

CCIE 52804

No port is blocked. R3 and R1 are pingable from R2.

It showed "neighbor IPv4 10.0.35.2 Unicast topology base removed from session BGP Notification received". Configured timer bgp and "maximum-path 2" on R5. The issue still exists.  

Review Cisco Networking for a $25 gift card