BGP and VRF

[adam.sellhorn]

I have a question regarding running BGP on a router with two VRF's:

Here is my config:

!

router bgp xxxxx

no synchronization

bgp log-neighbor-changes

network x.x.x.x mask 255.255.255.x

network x.x.x.x mask 255.255.255.x

network x.x.x.x mask 255.255.255.x

neighbor x.x.x.x remote-as xxxxx

no auto-summary

!

address-family ipv4 vrf VRF-2

redistribute connected

redistribute static

neighbor y.y.y.y remote-as xxxxx

neighbor y.y.y.y activate

no synchronization

exit-address-family

!

This works fine but now I need to add the following lines to the Global-VRF:

neighbor x.x.x.x prefix-list ROUTES_ALLOWED_IN in

neighbor x.x.x.x route-map PREPEND_RM out

When I do this the changes apply to the Global-VRF and VRF-2.

Should the Global-VRF be in it's own address-family? I think this would work but wanted to run it by you folks since this is my first deployment of VRF's and BGP.

Any help would be greatly appreciated.

Thanks.

[Edison Ortiz]

I'm not sure the behavior you are seeing. I applied your configuration here and it remained in the global portion of the BGP configuration. It didn't duplicate the command into the VRF-2.

What exactly are you seeing in your config?

Also, you only have one VRF, the global routing table isn't considered a VRF.

__

Edison.

[adam.sellhorn]

Thank you for your reply Edison. The problem I was having with the prefix-list and route-map in the global bgp configuration was that it was filtering routes in VRF-2. I have since created a second VRF and configured it as below to get it to work. I am hoping tomorrow to set up the previous config in my lab tomorrow and play around with it.

router bgp xxxxx

no synchronization

bgp router-id xxx.xxx.58.150

bgp log-neighbor-changes

no auto-summary

!

address-family ipv4 vrf VRF-1

redistribute connected

redistribute static

neighbor xxx.xxx.242.169 remote-as xxxxx

neighbor xxx.xxx.242.169 activate

no synchronization

exit-address-family

!

address-family ipv4 vrf Local

neighbor xxx.xxx.249.205 remote-as xxxxx

neighbor xxx.xxx.249.205 activate

neighbor xxx.xxx.249.205 prefix-list ROUTES_ALLOWED_IN in

neighbor xxx.xxx.249.205 route-map PREPEND_RM out

no synchronization

network xxx.xxx.58.150 mask 255.255.255.255

network xxx.xxx.58.151 mask 255.255.255.255

network xxx.xxx.129.0 mask 255.255.255.0

network xxx.xxx.129.32 mask 255.255.255.240

exit-address-family

[Edison Ortiz]

Good luck and let us know.

__

Edison.

[adam.sellhorn]

Well now that I have this setup in a lab environment it works fine. You were right Edison. There must have been something that I was overlooking when this was in production. Thanks for looking into this and sorry for wasting your time.

[Mohamed Sobair]

Hi,

just One note:

The Global IPV4 Bgp table is differ from the BGP VPNv4 table.

If you want to apply any Route filtering method you should know where u want it to be applied. As usually Normal BGP IPv4 carried out Between the PE-CE, And VPNv4 is carried out between PE-PE.

HTH

Mohamed

[adam.sellhorn]

I agree Mohamed and thank you for your reply. But in this case the provider terminated two VRFs to me via one T3 seperated by DLCIs so I had to use BGP IPv4 and VPNv4 on the CE.

[Mohamed Sobair]

Hi Adam,

Ok, so you are having Internet Service along with a VPN through the same MPLS-VPN Backbone provider.

Configuring per vrf instance on the bgp process shouldnt affect your Global BGP IPv4 config at all, pls double check and update.

HTH

Mohamed