Wouldn't an EIGRP distribution-list work here. I have a similar configuration and I just use a distribution-list to block the tunnel iterface subnet from being advertised back out the tunnel interface.
________________________________ From: Keller Giacomarro <[email protected]> To: robert shepherd <[email protected]> Cc: "[email protected]" <[email protected]> Sent: Wednesday, August 15, 2012 1:26 AM Subject: Re: [OSL | CCIE_RS] Tunnel interfaces So, you're almost in an ISP/customer situation here. The ISP routers are running OSPF, and your customer routers are running EIGRP with their respective parts of the network. Based on your question, it sounds like you would be doing EIGRP across the tunnel, correct? And you also have other EIGRP routes behind both R1 and R2 that I assume you want to share between R1 and R2. However, you shouldn't have a recursive routing problem in this scenario if you don't activate EIGRP on the physical interface being used for the tunnel. Recursive routing only applies to the tunnel endpoints. Unless there's a design requirement that says otherwise, you could advertise the physical interface only in OSPF, and the tunnel interface in EIGRP (and OSPF too, if desired). You would just need to be sure that the physical interface doesn't somehow end up in EIGRP (redistribution, etc). If there IS a design requirement to put the physical interface into EIGRP, you'll need to do some sort of route filtering to prevent R2 from learning R1's physical interface route via the tunnel. Distance is one option -- you could universally make EIGRP's AD higher than OSPF's on both routers. Your suggestion of 255, however, would make EIGRP unreachable and would cause no EIGRP routes to be entered into either routing table! You could also do an AD change on a specific set of prefixes, and could set the physical interface prefixes to have AD 255 or something higher than OSPF. Personally, I'd just put a distribute-list filter on both routers preventing them from learning the physical interfaces via EIGRP. Disclaimer: I'm still studying, so some of the above may be partially or fully wrong! Please consult your friendly neighborhood CCIE before taking anything that I say as fact. Keller Giacomarro [email protected] On Tue, Aug 14, 2012 at 10:36 PM, robert shepherd <[email protected]> wrote: > I know that we cannot learn the endpoint of a tunnel through the tunnel. > Let say we have R1 and R2 on different ends of the network running > eigrp. The rest of the network is running ospf. By default we will have > recursive routing happening due to the default AD's. In a case like this > would it be ok (if no restrictions are given) to always munipulate the AD > of eigrp to a distance of 255? Doing this would make sure that the end > point of the tunnel will never learn about itself through the tunnel. Is my > logic here correct? > _______________________________________________ > For more information regarding industry leading CCIE Lab training, please > visit http://www.ipexpert.com/ > > Are you a CCNP or CCIE and looking for a job? Check out > http://www.platinumplacement.com/ > > http://onlinestudylist.com/mailman/listinfo/ccie_rs _______________________________________________ For more information regarding industry leading CCIE Lab training, please visit www.ipexpert.com Are you a CCNP or CCIE and looking for a job? Check out www.PlatinumPlacement.com http://onlinestudylist.com/mailman/listinfo/ccie_rs _______________________________________________ For more information regarding industry leading CCIE Lab training, please visit www.ipexpert.com Are you a CCNP or CCIE and looking for a job? Check out www.PlatinumPlacement.com http://onlinestudylist.com/mailman/listinfo/ccie_rs
