I’m in the process of moving some of my routers from pfsense to vyos,
I have a router that participates in OSPF (single area) and private bgp sessions over VTI’s.
OSPF is setup and receiving routes from the rest of the area, as well as redistributing the routes learned from BGP without issue.
However the external (E2) routes received from the OSPF area are showing as ‘inactive’ and are not appearing in the ‘show ip route forward’ output. I assume this is also the reason why they (E2 routes) are not being redistributed to the BGP peer even though the OSPF network routing table is.
O>* 10.10.100.0/30 [110/30] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.8/30 [110/20] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.12/30 [110/30] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.16/30 [110/20] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.20/30 [110/40] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.33/32 [110/10] via 10.10.100.33, eth1, 00:28:38
O>* 10.10.100.35/32 [110/10] via 10.10.100.35, eth1, 00:28:33
O>* 10.10.100.37/32 [110/10] via 10.10.100.37, eth1, 00:28:33
O>* 10.10.100.38/32 [110/0] is directly connected, eth1, 00:28:40
O 172.30.40.0/24 [110/20] via 10.10.100.33, eth1 inactive, 00:21:32
O 172.30.41.0/24 [110/20] via 10.10.100.33, eth1 inactive, 00:21:32
O 172.30.42.0/24 [110/20] via 10.10.100.33, eth1 inactive, 00:21:32
O 172.30.43.0/24 [110/20] via 10.10.100.33, eth1 inactive, 00:21:32
O 172.30.44.0/24 [110/20] via 10.10.100.35, eth1 inactive, 00:21:27
O 172.30.47.0/24 [110/20] via 10.10.100.33, eth1 inactive, 00:21:32
O 172.30.48.0/28 [110/20] via 10.10.100.37, eth1 inactive, 00:21:27
O 172.30.49.0/24 [110/20] via 10.10.100.35, eth1 inactive, 00:21:27
show ip route forward output (public IP’s removed)
default via xxx.xxx.xxx.xxx dev eth0 proto static metric 20
10.10.100.0/30 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.8/30 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.12/30 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.16/30 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.20/30 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.32/29 dev eth1 proto kernel scope link src 10.10.100.38
10.10.100.33 via 10.10.100.33 dev eth1 proto ospf metric 20
10.10.100.35 via 10.10.100.35 dev eth1 proto ospf metric 20
10.10.100.37 via 10.10.100.37 dev eth1 proto ospf metric 20
10.10.100.38 dev eth1 proto ospf metric 20
10.10.100.68/30 dev vti0 proto kernel scope link src 10.10.100.69
10.10.100.200/30 dev eth2 proto kernel scope link src 10.10.100.201 linkdown
xxx.xxx.xxx.xxx/xx dev eth0 proto kernel scope link src xxx.xxx.xxx.xxx
Router C export routes via bgp to router A.
Router A can sees these routes and redistribute it to Router B via OSPF.
Router B sees these routes as recursive?
Do I understand the diagram correctly?
Router A & B are in OSPF area 0.0.0.0 with a handful of other routers - works ok
Router C exports routes via BGP to Router A - works ok
Router A redistributes learned BGP routes to OSPF - works ok
The problem comes when Router A doesn’t redistribute the OSPF routing table to Router C via BGP.
It redistributes the OSPF area routing table (10.10.100.0/24 network), but none of the OSPF E2 external routes are redistributed.
But after trying to get this topology to work for the better part of a day, I ended up redesigning the network slightly and segregating OSPF into different area’s with 0.0.0.2 as a ‘border’ area with extra routers D&E for a path redundancy.
