EIGRP routes prefered over eBGP

w.frye · ‎03-10-2010

Hi - I have a issue on a router where subnets learnt by EIGRP are getting put into the routing table over eBGP routes. I have not modified the default distance of BGP. I do not understand why this is the case. eBGP with admin distance 20 should be more preferable than EIGRP (external routes 170).

The only other thing is that this router is doing redistribution both ways (EIGRP into BGP and BGP into EIGRP). The router is learning the same subnets via both routing protocols (I am required to redistibute both ways in our scenario). I have shown the routing table and a show ip bgp nei recieved-rpoutes below. The networks I want to see get replaced by BGP are 192.168.130.0/24 - 192.168.133.0/24. (I have highlighted in red below).

How can I resolve this?

F#sh ip route

Codes: 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

Gateway of last resort is not set

D EX 192.168.132.0/24
[170/2844160] via 192.168.10.2, 00:10:14, FastEthernet0/1
D EX 192.168.133.0/24
[170/2844160] via 192.168.10.2, 00:10:14, FastEthernet0/1

B 192.168.8.0/24 [20/0] via 192.168.9.1, 00:03:37

C 192.168.9.0/24 is directly connected, FastEthernet0/0

C 192.168.10.0/24 is directly connected, FastEthernet0/1

D EX 192.168.130.0/24

[170/2844160] via 192.168.10.2, 00:10:14, FastEthernet0/1

D EX 192.168.131.0/24

[170/2844160] via 192.168.10.2, 00:10:14, FastEthernet0/1

B 192.168.5.0/24 [20/1000] via 192.168.9.1, 00:03:38

B 192.168.6.0/24 [20/1000] via 192.168.9.1, 00:03:38

D 192.168.7.0/24 [90/284160] via 192.168.10.2, 1d01h, FastEthernet0/1

B 192.168.128.0/21 [20/1000] via 192.168.9.1, 00:03:39

B 192.168.0.0/20 [20/1000] via 192.168.9.1, 00:03:39

F#sh ip bgp neighbors 192.168.9.1 received-routes

BGP table version is 70, local router ID is 192.168.10.1

Status codes: s suppressed, d damped, h history, * valid, > best, i - internal,

r RIB-failure, S Stale

Origin codes: i - IGP, e - EGP, ? - incomplete

Network Next Hop Metric LocPrf Weight Path

*> 192.168.0.0/20 192.168.9.1 1000 0 65201 ?

*> 192.168.5.0 192.168.9.1 1000 0 65201 ?

*> 192.168.6.0 192.168.9.1 1000 0 65201 ?

*> 192.168.8.0 192.168.9.1 0 0 65201 ?

*> 192.168.128.0/21 192.168.9.1 1000 0 65201 ?

* 192.168.130.0 192.168.9.1 1000 0 65201 ?

* 192.168.131.0 192.168.9.1 1000 0 65201 ?

* 192.168.132.0 192.168.9.1 1000 0 65201 ?

* 192.168.133.0 192.168.9.1 1000 0 65201 ?

Giuseppe Larosa · ‎03-10-2010

Hello W.Frye,

this happens because of mutual redistriibution and because AD is not considered in the BGP table.

you can check what happens for a single IP prefix with

sh ip bgp

you should see a locally generated BGP route with IP next-hop 0.0.0.0 and weight 32,768 (default weight for locally generated prefixes)

as best path, you may receive the BGP prefixes also from eBGP peer 192.168.9.1 but with weight 1000 < 32,768

* 192.168.130.0 192.168.9.1 1000 0 65201 ?

* 192.168.131.0 192.168.9.1 1000 0 65201 ?

* 192.168.132.0 192.168.9.1 1000 0 65201 ?

* 192.168.133.0 192.168.9.1 1000 0 65201 ?

you can see that the '>' symbol is missing meaning that these are not considered best paths.

so this router is trying to advertise its locally generated IP prefixes into BGP.

If you want eBGP routes to be preferred you should give them a weight higher then 32,768 for example 40,000

if you do so you should be able to see eBGP routes installed in BGP table and then in the IP routing table.

Note:

if you perform mutual redistribution in multiple nodes you should take care of unwanted routing feedback.

Hope to help

Giuseppe

naveen_b81 · ‎03-10-2010

H Frye,

Can you please post more information preferably in a diagram as to where these networks 192.168.132.0/24 are placed and how are they getting introduced into EIGRP and BGP?

Best Regards,

Naveen

jcozzupoli · ‎03-10-2010

Hi,

You should be weary of running into a race-condition in your routing table with both EIGRP and BGP "racing" each other as to what is prefered with each protocol updating the routing table constantly, this will cause huge issues with CPU flat lining at 100%!

You should really only be advertising your localky learned routes via redistribution and not externally learnt routes using a distribute-list for example.

Also, yes could you please share with us a Layer 3 network diagram

Regards,

Joe.