cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1189
Views
0
Helpful
14
Replies

EIGRP External/Internal routes, need solution!

glenthms
Level 1
Level 1

I have attached a visual depiction of what we have setup. Our Site B has two paths to get to our Core for Data services. The primary path is over a T1 via MPLS to our Site A which then has a link to our core over a seperate VPN tunnel using GRE. Our issue is this. Site B networks are seen in our Core as Internal routes and are preferred over the External routes that Site A advertises into the core. Site B advertises its own network to our core over its backup tunnel and Site A advertises Site B networks over its own VPN tunnel. My question is this, if were running EIGRP 10 on all of our routers and also with the provider in MPLS, how can I change how routes appear to the core from Site B path to also be External routes? The only solution we can come up with is to use two instances of EIGRP and redistribute between each which changes the routes advertised from Site B into the core as External routes and then we get symetrical routing. Does this seem like the only solution? Help!

1 Accepted Solution

Accepted Solutions

Hello Glen,

here it is a link:

http://www.cisco.com/en/US/docs/ios/12_2t/12_2t15/feature/guide/fteipece.html

the trick is that PE speaks EIGRP with your CE and transports EIGRP routes within MP BGP extended communities

Hope to help

Giuseppe

View solution in original post

14 Replies 14

Jerry Ye
Cisco Employee
Cisco Employee

Personally, I do not prefer to create two EIGRP AS and redistribute between them because a small mistake can create routing loop for your network.

Just a thought, what is your address block for Site B? Are they in a consecutive block? The reason I am asking is because I am thinking to use EIGRP summary routes to create a fix.

Remember how a route gets into the routing table:

1) more specific routes are always prefer

2) lowest admin distance if the prefix are the same

3) lowest metric if admin distance are the same

HTH,

jerry

Forgot one thing, please lab it before put it in production.

jerry

Thanks Jerry, the address block for Site B is really small, half a class C. The core only permits that specific network in as well.

Hm... is it possible to create /23 summary in EIGRP and allow that in your core?

Just an example:

1) Let's say 10.2.2.0/25 is your address block for Site B.

2) If 10.2.2.128-10.2.3.255 are not in use.

3) You can create a summary of 10.2.2.0/23.

This way, the router will prefer the MPLS BGP because it sees a more specific prefix and your internal with a summary (less specific).

HTH,

jerry

Hello Glen,

Or in your core router, under the specific eigrp process, you can simply change the admnistrative distances for internal and external routes like following

Core1(config)#router eigrp 10

Core1(config-router)#distance eigrp 90 89

Now external route will be installed to route table since it has lower AD. Keep in mind that this command will only effect eigrp process 10, so if you have other eigrp processes in core routers for other sites, they will stay inact with default ADs.

Regards

Unfortunately we have a lot of customers that use EIGRP 10 and this would create an issue with them. Thanks though!

Jerry thanks again, the other /25 of the subnet is already routed in the core, so I cannot create a summary route.

"Unfortunately we have a lot of customers that use EIGRP 10 and this would create an issue with them. Thanks though!"

Distance command can also be used "per route" , but as I recall, it can not be used for EX routes. So what you can do is, in Core router 1 that should install D EX into its routing table instead D for that specific route only and not affect the rest, issue the following commands

access-list 10 permit y.y.y.y z.z.z.z

Core1(config-router)#distance 171 x.x.x.x w.w.w.w 10

y.y.y.y=Subnet of site B

z.z.z.z=Wildcard mask of that subnet

x.x.x.x=Ip address(subnet) of internal EIGRP neighbor(s) that advertises the subnet of site B as internal route and causes core router 1 to prefer it instead D EX

w.w.w.w=Wildcard mask for neighbor IP(or subnet)

Ill run it past our core engineers. Good idea!

Hello Glen,

How did you end up in this one?

Regards

Hello,

We were going to attempt this as a solution in our core however this distacne is not propagted correctly up to further upstream routers. Per cisco.

Administrative distance is generally used to determine the method by which a route was learned. If set

incorrectly, the individual router cannot choose a redistributed route instead of the actual best path.

Administrative distance is not propagated to other routers. Routing protocols rely on the fact that all

routers choose the same path given the same set of parameters. Altering parameters on a single router

can lead to routing loops.

Hello Glen,

the AD is not part of any routing protocol advertisement so what you see is correct behavior: AD settings have a single node scope you need to propagate the commands on all the involved routers.

The best solution in your case should be that of your MPLS ISP to support an EIGRP PE-CE protocol communication.

Hope to help

Giuseppe

Giuseppe thanks for your response. Do you have more information about EIGRP PE-CE protocol communication?

Hello Glen,

here it is a link:

http://www.cisco.com/en/US/docs/ios/12_2t/12_2t15/feature/guide/fteipece.html

the trick is that PE speaks EIGRP with your CE and transports EIGRP routes within MP BGP extended communities

Hope to help

Giuseppe

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Innovations in Cisco Full Stack Observability - A new webinar from Cisco