Redistributing loop ?

Jules Sanchez · ‎12-12-2014

Hi all,

I'm scratching my head at this one, I'm sure it's something simple though.
The network is completely dynamic, very standard MPLS Layer3 VPNs etc..

The network 10.85.2.0/24 has been disconnected.
Yet Router1 learns it via BGP from Router2.
And Router2 learns it via RIP from Router1

Any assistance is greatly appreciated.

Router 1

core1#show ip route vrf MGT

Routing Table: MGT
Codes: L - local, C - connected, S - static, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2
i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, * - candidate default, U - per-user static route
o - ODR, P - periodic downloaded static route, H - NHRP, l - LISP
a - application route
+ - replicated route, % - next hop override

Gateway of last resort is 10.85.1.254 to network 0.0.0.0

S* 0.0.0.0/0 [1/0] via 10.85.1.254
10.0.0.0/8 is variably subnetted, 15 subnets, 3 masks
B 10.75.1.0/24 [200/0] via 10.8.135.23, 7w0d
B 10.75.3.0/24 [200/0] via 10.8.134.23, 7w0d
C 10.85.1.0/24 is directly connected, GigabitEthernet0/0/1
L 10.85.1.3/32 is directly connected, GigabitEthernet0/0/1
B 10.85.2.0/24 [200/1] via 10.8.132.24, 04:25:34
B 10.85.3.0/24 [200/0] via 10.8.133.23, 7w0d
C 10.95.1.100/32 is directly connected, Loopback1500
R 10.95.1.101/32 [120/1] via 10.85.1.4, 00:00:14, GigabitEthernet0/0/1
B 10.95.1.102/32 [200/0] via 10.8.135.23, 7w0d
B 10.95.1.103/32 [200/0] via 10.8.135.24, 7w0d
B 10.95.1.104/32 [200/0] via 10.8.134.23, 7w0d
B 10.95.1.105/32 [200/0] via 10.8.134.24, 7w0d
B 10.95.1.106/32 [200/0] via 10.8.133.24, 7w0d
B 10.95.1.107/32 [200/0] via 10.8.133.23, 7w0d
B 10.95.1.114/32 [200/1] via 10.8.132.24, 04:25:34
S 192.168.50.0/24 [1/0] via 10.85.1.254

Router 2

core2#show ip route vrf MGT

Routing Table: MGT
Codes: L - local, C - connected, S - static, R - RIP, M - mobile, B - BGP
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2
i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, * - candidate default, U - per-user static route
o - ODR, P - periodic downloaded static route, H - NHRP, l - LISP
+ - replicated route, % - next hop override

Gateway of last resort is 10.8.132.23 to network 0.0.0.0

B* 0.0.0.0/0 [200/0] via 10.8.132.23, 7w0d
10.0.0.0/8 is variably subnetted, 15 subnets, 2 masks
R 10.75.1.0/24 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.75.3.0/24 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
C 10.85.1.0/24 is directly connected, GigabitEthernet0/1
L 10.85.1.4/32 is directly connected, GigabitEthernet0/1
R 10.85.2.0/24 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.85.3.0/24 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.100/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
C 10.95.1.101/32 is directly connected, Loopback1500
R 10.95.1.102/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.103/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.104/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.105/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.106/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.107/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
R 10.95.1.114/32 [120/1] via 10.85.1.3, 00:00:14, GigabitEthernet0/1
S 192.168.50.0/24 [1/0] via 10.85.1.254

Jon Marshall · ‎12-12-2014

It's difficult to say without knowing the full topology.

When you say it has been disconnected do you mean the interface has been shut down ?

Are you doing mutual redistribution on both routers ?

If I'm reading the routing tables correctly core1 seems to be learning it from a BGP neighbour and then redistributes this into RIP and advertises it to core2 ie. core1 and core2 are connected via the 10.85.1.x subnet.

But core2 is not learning the route from core1 as far as I can see. It is learning it from 10.8.132.24, whatever device that is ?

Jon

Jules Sanchez · ‎12-16-2014

These are loopbacks.

Router 1 = 10.8.132.23

Router 2 = 10.8.132.24

My biggest concern is that even though that network no longer existed, 10.85.1.2.
It was still in the routing table.
It almost looked like this network was continuously being learn from neighbouring routers even though the network was no loner connected.

Router 2 would normally learn this network from a direct link from another DataCenter via BGP. Then redistribute it.

The topology is quite simple, 5 sites, all with dual routers acting as P/PE in a standard MPLS configuration.

All Router are full mesh with each other via BGP.

milan.kulik · ‎12-17-2014

Hi,

the topology is not clear.

Are both routers on the same site?

Why are they peering via RIP additionally to BGP?

Have you tried to break the redistribution loop by prefix tagging?

Best regards,

Milan

milan.kulik · ‎12-15-2014

Hi, so Router1 learns the 10.85.2.0/24 prefix via BGP from Router2. And is rediostributing the prefix to RIP? While Router2 learns the same prefix via RIP (from Router1) and redistributes it to BGP? While in fact the 10.85.2.0/24 subnet is Down? To break this redistribution loop, you could e.g.: 1) to add some tag to the RIP prefixes when redistributing them from BGP to RIP. 2) And on the other router to create a route-map denying prefixes with that specific tag to be redistributed to BGP from RIP. Best regards, Milan