cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
827
Views
0
Helpful
1
Replies

leaking subscribers between VRFs

mike-stevenson
Level 1
Level 1

Hi,

I have two VRFs, lets call them internet, and customers_1.

PPPoE, and IPoE subscribers terminate in the customers_1 VRF, I am wanting to leak these addresses into the internet vrf.

I have configured the relevant import and export statements, and see the routes, however the routes are advertised into the internet vrf with a next hop of 0.0.0.0 thus, they do not appear in CEF (seen via sh cef vrf interent), and traffic is not forwarded.

Can anyone let me know how I would achieve this.

here is the customer_1 vrf routing table (partial):

B    3.3.3.0/24 [200/0] via 202.74.33.249 (nexthop in vrf internet), 00:00:08

B    4.4.4.0/24 [200/0] via 202.74.33.249 (nexthop in vrf internet), 00:00:08

S    100.64.0.0/12 is directly connected, 01:11:20, Null0

C    100.64.0.0/17 is directly connected, 1w3d, Loopback2

L    100.64.0.1/32 is directly connected, 1w3d, Loopback2

A   100.64.0.99/32 is directly connected, 1w1d, Bundle-Ether100.1.ip8

B    103.241.56.0/22 [200/0] via 202.74.33.249 (nexthop in vrf internet), 00:00:08

A    116.251.122.4/32 is directly connected, 1w1d, Bundle-Ether100.1.pppoe5

A    116.251.193.254/32 is directly connected, 1w1d, Bundle-Ether100.1.ip7

here is the internet vrf rotuing table (partial)

B    3.3.3.0/24 [200/0] via 202.74.33.249, 00:00:02

B    4.4.4.0/24 [200/0] via 202.74.33.249, 00:00:02

B    100.64.0.99/32 [200/0] via 0.0.0.0 (nexthop in vrf customers_1), 00:47:47, Bundle-Ether100.1.ip8

B    116.251.122.4/32 [200/0] via 0.0.0.0 (nexthop in vrf customers_1), 00:47:47, Bundle-Ether100.1.pppoe5

B    116.251.128.0/18 [200/0] via 202.74.33.249, 00:00:02

B    116.251.192.0/21 [200/0] via 202.74.33.249, 00:00:02

B    116.251.193.254/32 [200/0] via 0.0.0.0 (nexthop in vrf customers_1), 00:47:47, Bundle-Ether100.1.ip7

B    202.74.33.58/32 is directly connected, 00:44:45, Loopback1 (nexthop in vrf customers_1)

the follwoing  are the vrf export/import statements - i have no route-maps yet, as am just trying to get basic connectivity going first.

vrf internet

address-family ipv4 unicast

  import route-target

   65536:200

   65536:100

  !

  export route-target

   65536:200

!

vrf customers_1

address-family ipv4 unicast

  import route-target

   65536:200

   65536:100

  !

  export route-target

   65536:100

in BGP i have for the customers vrf:

vrf customers_1

  rd 65536:100

  address-family ipv4 unicast

   redistribute ospf customers_1 match internal external

   redistribute subscriber

  !

many thanks,

mike

1 Reply 1

Rivalino Tamaela
Cisco Employee
Cisco Employee

Mike,

I do not have other alternative solutions, the only one i can suggest is by using Access-list forwarding (ABF) on incoming interface from internet. The good thing about this is that you can have one ACE for pool of your customer's ip address. Put summarized destination address and point to the VRF as the destination. It is called ABF VRF select. This cause the ingress packet from internet with destination of vrf customers_1 ip address will be forwarded using VRF customers_1.

regards,

rivalino