Two dynamic protocols routing with same AD

[Haval.Darwesh]

I have learned what is AD in routing protocols, and how routers are using them for preferring a routing protocol over another.

my question is, what happens if I configured two dynamic routing protocols (for example EIGRP and RIP) on a network but changing AD of both protocols to the same value (for example changing AD of RIP to 90 which is equal to EIGRP)?!

I have test it on a simple network, I have configured both RIP and EIGRP on two routers, and then I changed AD of RIP to 90 to be equal to EIGRP AD. I found that both routers are going to use RIP?

Why?!

[MHM Cisco World]

If AD same then tie break is metric'

Check metric of both eigrp and rip

[Flavio Miranda]

Hi

 The test you did does not make sense. There is a whole engeneering behind developing this concept  and implementing it in the software expecting us to follow the rules. The result you got probably only the vendor is able to explain and chances are the IOS choose it randamly or it choose RIP because it is older and therefore lighter from memory and CPU perspective but this simple guessing. 

[David Ruess]

Hello,

Yeah I'd have to agree with @MHM Cisco World as the metric determines the route after AD. So RIPs metrics are in hop counts and EIGRP uses Bandwidth and delay. So if a RIP router has a hop count of 2...its metrics is probably....2. Whereas EIGRP will never be that low. Even at the max for RIP of 15 hops I don't think you can get EIGRP metrics that low to be under 15 (maybe). The metrics fo rthe 2 routing protocols are not really comparable so its difficult to say without more testing and observation. Depending on the platform you can do a couple things. You can run the command debug ip routing which I believe debugs changes to the routing table. So you would set up EIGRP first. and then RIP. Then once you manipulate RIP to have the AD of 90 then you should be able to see the debug messages of its "decision" to use RIP over EIGRP with same AD.

You can also look at the show ip eigrp events command and it shows the last 500? I think messages EIGRP went through to see why it removed the EIGRP route out of the routing table. If we are correct you should see something like removed due to better metric/route or something like that.

-David

[MHM Cisco World]

Additional to @David Ruess  

That way when redistrubte one protocol to other try match metric'

The eigrp use metric in K (1000) value where opsf and  bgp use metric with hundreds value.

Thanks 

MHM

[Joseph W. Doherty]

I believe @MHM Cisco World and @David Ruess posts explain the reason, i.e. when RIP and EIGRP have same AD, for same prefix, RIP's metric is better.

Possibly an even faster way to see this is look at a route(s) from show IP route and look for two numbers within braces, first value should be AD, second value metric.  E.g. #.#.#.#/# [AD#/metric#] via . . .

[Martin L]

is this in PT? How did u test this? what gear ? what IOS versions ?

Regards, ML
**Please Rate All Helpful Responses **

[Joseph W. Doherty]

"I have learned what is AD in routing protocols, and how routers are using them for preferring a routing protocol over another."

BTW, AD is not "in" routing protocols, it's a Cisco proprietary methodology that's used to possibly tie break when the same network prefix is received from different kinds of sources.

[Martin L]

In your case of manually configuring AD values on a single router to have the same AD value in 2 different protocols, Cisco router should ignore your manual AD configuration (ignore configured AD value) and select protocol based on default AD values. The lower AD wins, aka 90 vs 120. Also, ADs are locally significant to a particular device; meaning are not advertised. 

So, either you made mistake somewhere or you are using simulator and not real IOS routers.  Please provide details about your lab gear/software and settings/configurations.

This has been discussed in depth recently.  see https://community.cisco.com/t5/routing/routing-table-routes-with-different-administrative-distance/td-p/4908572

Regards, ML
**Please Rate All Helpful Responses **