Default routes into EIGRP

[richardmcmahon]

We have the following setup

-----Internet----

| |

BGP BGP

Router1=========Router2

| |

| |

| |

| |

DC1 DC1

Router1=========Router2

Ok dodgy diagram asside, we run BGP and EIGRP on the BGP routers. A static default route with a high metric is redistributed from both BGP routers to the DC routers which only run EIGRP. This works fine and for all other static routes coming from the BGP routers for other internal networks show two routes available for load balancing. The defalt though only shows as a single route. If I disable eigrp on the router which is shown as the default then it times out and the other takes over. I would ideally like to load balance the traffic on both uplinks. I have tried forcing the metric on both BGP routers to be the same for their routes but it appears to make no difference. Any help gratefully appreciated (and rated) :)

Thanks,

Richard

[Edison Ortiz]

Richard,

Let's see 'show ip eigrp ne' and 'show ip protocols | be eigrp' from the 4 routers.

Thanks

[royalblues]

Richard,

Can you post the configs of the 4 routers

Narayan

[richardmcmahon]

Hopefully this is enough.

Thanks,

Richard

[Edison Ortiz]

Richard,

Can I also see a 'show ip route 0.0.0.0' from both DC routers ?

You can check out this link for traffic share-balance under EIGRP and see if it helps

http://www.cisco.com/warp/public/103/19.html#traffic

[richardmcmahon]

dc1

Routing entry for 0.0.0.0/0, supernet

Known via "eigrp 20", distance 170, metric 2816, candidate default path, type external

Redistributing via eigrp 20

Last update from 10.1.64.32 on Vlan201, 4d22h ago

Routing Descriptor Blocks:

* 10.1.64.32, from 10.1.64.32, 4d22h ago, via Vlan201

Route metric is 2816, traffic share count is 1

Total delay is 10 microseconds, minimum bandwidth is 1000000 Kbit

Reliability 255/255, minimum MTU 1500 bytes

Loading 1/255, Hops 1

dc2

Routing entry for 0.0.0.0/0, supernet

Known via "eigrp 20", distance 170, metric 2816, candidate default path, type external

Redistributing via eigrp 20

Last update from 10.1.64.32 on Vlan201, 4d22h ago

Routing Descriptor Blocks:

* 10.1.64.32, from 10.1.64.32, 4d22h ago, via Vlan201

Route metric is 2816, traffic share count is 1

Total delay is 10 microseconds, minimum bandwidth is 1000000 Kbit

Reliability 255/255, minimum MTU 1500 bytes

Loading 1/255, Hops 1

[Wilson Samuel]

Hi,

I guess you are trying to use Load Balancing / Load Sharing between the 2 EIGRP Routers (DCs) and the BGP Routers, please correct me if I'm wrong.

If that is the case, it could be done with the help of adjusting the EIGRP Varaince.

May I request you to post the sh ip route from both the DC routers, and based on that you may set the variance by 2 and it should do the needful.

Regards,

Wilson Samuel

[richardmcmahon]

Hi,

That is exactly what I am trying to achieve. The problem is that default route wont load balance. The other routes coming from the BGP routers are in there twice for load sharing but not the default. See full ip route below from one of the dc routers

10.0.0.0/8 is variably subnetted, 26 subnets, 6 masks

D EX 10.2.196.96/27 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.196.64/27 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D 10.2.195.64/26 [90/3072] via 10.1.64.32, 1w0d, Vlan201

[90/3072] via 10.1.64.31, 1w0d, Vlan201

D EX 10.2.196.32/27 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.202.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.200.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.201.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.198.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.199.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.196.0/27 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.197.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D EX 10.2.194.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D 10.2.195.0/26 [90/3072] via 10.1.64.32, 1w0d, Vlan201

[90/3072] via 10.1.64.31, 1w0d, Vlan201

D 10.2.192.0/25 [90/3072] via 10.1.64.32, 1w0d, Vlan201

[90/3072] via 10.1.64.31, 1w0d, Vlan201

D EX 10.2.193.0/24 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D 10.2.196.224/27 [90/3072] via 10.1.64.32, 1w0d, Vlan201

[90/3072] via 10.1.64.31, 1w0d, Vlan201

D EX 10.2.196.128/27 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D 10.2.195.128/25 [90/3072] via 10.1.64.32, 1w0d, Vlan201

[90/3072] via 10.1.64.31, 1w0d, Vlan201

D EX 10.2.192.128/25 [170/3072] via 10.1.64.32, 4d22h, Vlan201

[170/3072] via 10.1.64.31, 4d22h, Vlan201

D*EX 0.0.0.0/0 [170/2816] via 10.1.64.32, 4d22h, Vlan201

[mark.edwards]

Hi Peter, You need to start by looking at the EIGRP topology table. Firstly you must see 2 entries in the topology table for the two default routes. If you don't see this you may not be meeting the "feasibility condition" for EIGRP and you may have to alter metric to meet this (Feasible distance is the best metric along a path to a destination network, including the metric to the neighbour advertising that path. Reported distance is the total metric along a path to a destination network as advertised by an upstream neighbour. A feasible successor is a path whose reported distance is less than the feasible distance ).

Can you post the topology table of the DC routers? Also I notice when you redistribute static under EIGRP you do not set a default distance. This should also be set or static?s won't get redistributed.

[richardmcmahon]

Hi Mark,

Attached below is an excert from the eigrp top table. As you can see I have two routes for other routes distributed from the bgp routers. These are also redistributed statics so missing the the default distance doesnt seem to be preventing them from being redistributed. Out of interest what is the recommended value for the default distance.

Thanks,

Richard

DC1

P 0.0.0.0/0, 1 successors, FD is 2816

via 10.1.64.32, Vlan201

P 10.2.196.96/27, 2 successors, FD is 3072

via 10.1.64.31, Vlan201, serno 840

via 10.1.64.32, Vlan201

P 10.2.196.64/27, 2 successors, FD is 3072

via 10.1.64.31, Vlan201, serno 841

via 10.1.64.32, Vlan201

P 10.2.195.64/26, 2 successors, FD is 3072

via 10.1.64.31, Vlan201

via 10.1.64.32, Vlan201

DC2

P 0.0.0.0/0, 1 successors, FD is 2816

via 10.1.64.32, Vlan201

P 10.2.196.96/27, 2 successors, FD is 3072

via 10.1.64.31, Vlan201, serno 919

via 10.1.64.32, Vlan201

P 10.2.196.64/27, 2 successors, FD is 3072

via 10.1.64.31, Vlan201, serno 920

via 10.1.64.32, Vlan201

P 10.2.195.64/26, 2 successors, FD is 3072

via 10.1.64.31, Vlan201

via 10.1.64.32, Vlan201

[mark.edwards]

Hi Richard. Not sure how the statics are being redistributed without a default metric. Normally this is required but perhaps I don't understand the setup here and have missed something!

Can you do a show ip eigrp topology on the BGP routers and also show ip bgp.