[c-nsp] Adding connected routes in a VRF

[Ross Vandegrift]

On Mon, Dec 08, 2008 at 02:50:05PM -0600, Nick Griffin wrote:
> You have to manually add host routes as the next hop since you can't add the
> router itself, another solution I found that work was this:
> 
> "BGP Support for ipv4 Prefix Import". This for me worked well, you just need
> to make sure that the prefixes you wish bring in from the Global Table exist
> in the BGP GRT RIB, see example below:

No, this is a different feature.  Prefix import permits you to leak
traffic out of a VRF into the global table.  I don't want traffic to
move between VPNs.  Here's a bit more complete config that better
displays the difference:


ip vrf foobar
 rd 1:1
!
interface GigabitEthernet1/1
 ip address 10.0.100.1 255.255.255.0
!
interface GigabitEthernet1/2
 ip vrf forwarding foobar
 ip address 10.0.200.1 255.255.255.0
!
! second route throws an error
ip route 192.168.100.0 255.255.255.0 GigabitEthernet1/1
ip route vrf foobar 192.168.200.0 255.255.255.0 GigabitEthernet1/2


router#show ip route 192.168.100.0
Routing entry for 192.168.100.0/24
  Known via "static", distance 1, metric 0 (connected)
  Redistributing via ospf 10
  Advertised by ospf 10 subnets
  Routing Descriptor Blocks:
  * directly connected, via GigabitEthernet1/1
      Route metric is 0, traffic share count is 1


If I have two machines connected to Gi1/1, numbered in 10.0.100.0/24,
and I assign secondary addresses from 192.168.100.0/24, those host
addresses will work fine if I know that both servers have 10.0.100.1
as the next-hop for all of their routes.

While this probably isn't a common scenario, I have a few common
server scenarios where this helps a lot in converging networks onto
fewer routers.


Ross

-- 
Ross Vandegrift
ross at kallisti.us

"If the fight gets hot, the songs get hotter.  If the going gets tough,
the songs get tougher."
	--Woody Guthrie