Collecting as BGP RR client

Forward Enterprise has been successfully supporting the BGP Monitoring Protocol ( BMP) as the method of choice to quickly and efficiently extract extremely large (Internet size) IP tables. Unfortunately some critical Cisco routing platforms do not fully support this capability. On routers running IOS-XR software, BMP is available only for the default VRF, which makes it insufficient for Forward Enterprise to build a complete model of the device.

An alternative solution is to modify the BGP configuration of these devices by adding the Forward platform as " route-reflector-client". Forward Enterprise supports this additional option for Cisco IOS-XR devices

BGP configuration

This is the BGP configuration that needs to be added to Cisco IOS-XR devices carrying very large IP tables.
No configuration is needed on Forward Collector.
In this configuration, the user will enter the Forward Collector's IP address and configure it as Route Reflector client:

neighbor <Forward Collector IP address>
 remote-as <AS Number>
 address-family ipv4 unicast
   route-reflector-client
   route-policy allow_out out
 !
 address-family vpnv4 unicast
   route-reflector-client
   route-policy allow_out out
 !
 route-policy allow_out
   pass
 end-policy

tip

Forward Collector establishes the peering relationship only after it parses the BGP configuration and it realizes that the device has been configured to export its routes to Forward. The peering relationship is established only for the time needed to download the routes and it is terminated thereafter.

For a neighbor with IP address A.B.C.D whose BGP peering is done inside a non-default VRF, the following configuration is needed as well:

community-set <NEIGHBOR_A_B_C_D_COMMUNITY>
 <unique community string for neighbor A.B.C.D> (example: 123:123)
end-set

route-policy <MARK_BGP_NEIGHBOR_A_B_C_D>
 set community <NEIGHBOR_A_B_C_D_COMMUNITY> additive
 pass
end-policy

Then as part of the BGP config for that neighbor, add the following:

router bgp <AS Number>
 vrf <VRF name>
   neighbor <A.B.C.D>
     address-family ipv4 unicast
        route-policy <MARK_BGP_NEIGHBOR_A_B_C_D> in

BGP configuration​

BGP configuration