Troubleshooting Route Redistribution

the route redistribution issue we saw in our prior video had a single point it was router r2 where route redistribution was happening however if we were designing a network ourselves we might look at that redistribution point and think that that's a potential single point of failure instead we might want to have two or more routers that are redistributing between different autonomous systems different routing domains we can certainly do that but in some cases that might introduce a bit of a troubleshooting issue for example let's say that router two on screen wants to send traffic to the 10.1.1.10 t4 network and router Artoo's IP routing table says that the next hop is router r d1 which is a router that's doing redistribution r d1 sins the packets down to router r1 and interestingly router r1 sins the packets on two are d2 and RD two sends those packets back into our original autonomous system that could be suboptimal routing relieving our routing domain going through another routing domain another autonomous system and then coming back into our own now routing domain and you might run into a situation like this depending on how many domains you have how many boundary routers you have what specific routing protocols you're using how things are configured and in this video we want to talk mostly in theory about how we would troubleshoot an issue like this what we could do is set really really high metric values on a route that's being redistributed into a routing domain that way if I'm in autonomous system number two and I'm trying to send packets to a destination that also lives in autonomous system number two I'm probably not going to be leaving my autonomous system because the metric would be too high to go through another autonomous system it would be a much lower metric to stay within my own autonomous system something else we might do to combat a situation like this is to statically set the administrative distance for a routing protocol remember that rip has a default administrative distance of 120 OSPF has a default administrative distance of 110 ya GRP as a default administrative distance of I mean except for external routes that get injected into EIGRP those have an administrative distance of 170 in fact since ERP does distinguish between externally learned routes and routes learned within the autonomous system and it gives a higher administrative distance to those externally learned routes vig RP by itself does a great job of preventing a situation like this from happening but depending on what routing protocol you're using and how things are configured you might in some cases need to statically configure the administrative distance for a routing protocol so let's hop out to a live interface real quick and let's take a look at how easy it is to set the administrative distance for a routing protocol I'm sitting on a router Rd one and it is configured for route redistribution and if we do a show IP route we can see that we've learned routes via OSPF we've learned routes via EIGRP and notice the ADEs the administrative distances of these different routing protocols ERG RP has an ad of 90 that's the default OSPF has an ad of 110 that's the default however in some troubleshooting scenarios we might want to make OSPF more believable than eigrp here's how we could do that we could give the distance command and by the way that command works not just with OSPF it also works with eigrp it works with rip here's how we do it we go into router configuration mode for OSPF process ID 1 and I'm going to say distance and we give the new ad I'm gonna make it 80 I want to make it more believable than eigrp and now if I look at my IP routing table again look at this my OSPF routes now have a more believable administrative distance than ERP it's now 84 OSPF and you'll see that I don't have any eigrp learned routes in the IP routing table now because I've learned all of them via OSPF that's interesting but something important to realize is that this ad of 80 is only locally significant it only applies to router Rd one where I've configured it in other words this ad is not being advertised out to other routers this is not going to influence the routing decisions that other routers make this influences rd1 another way for us to overcome this potential routing loop is to filter routes as they're redistributed for example if we know all of the 172 dot 1600 / 16 networks reside in a routing domain one we might go into the routers or redistributing into that routing domain and filter that route out after all that route lives in that domain no sense advertising that route into the domain from another domain and you can use a route map to do that filtering another fairly clever way to overcome this routing loop issue is to tag your route as it's being redistributed from one routing domain into another and when I talk about the tags please realize we can use these for lots of different reasons not just for redistribution but what is a tag it's a value that we can assign to a route and a tag isn't really measuring anything it's not a specific unit of measure for bein with or delay or anything like that it's just a label that we put on a route let's check out an example of what we could do with tags what we could do is say that we want to assign a tag of 10 to routes being redistributed into autonomous system one on-screen and I've just drawn it for router Rd 1 but we would do the same thing for router Rd 2 but in addition to setting the tag to 10 for routes going into autonomous system 1 what if we did this also what if we said that we're going to deny any routes from being redistributed that had a tag of 10 in the example on screen we've got a route going from autonomous system 2 into autonomous system 1 but as that route gets redistributed it's given a tag of 10 and Rd 1 in this example is assigning that tag if that same route were to then try to come back into autonomous system 2 via Rd 2 that's not going to be allowed because Rd 2 is going to have a route map that says we're not going to allow the redistribution of any route that has a tag of 10 going into autonomous system 2 let's go out to a live interface now and take a look at how we can configure these tags we're here again on router Rd 1 and what we want to do is to say if we're redistributing a route for me I GRP into OSPF we want to give it a label we're going to give it a tag of 10 and we can do that with a route map let's go into global configuration mode and let's create a route map I'll say route - map tag at 10 I don't need to say permit and I don't need to give a sequence number because the default is permit and I'm only gonna give one instruction I'm gonna say set the tag to 10 and that's the first of our two route maps we need to create let's create a second route map I'm gonna say route - map this time it's gonna be deny tag 10 and I will say deny this time and I'll give it a sequence number what am i wanting to deny I'm wanting to deny routes that have a tag of 10 so I will say match tag 10 and we're going to deny those from being redistributed from OSPF back into EIGRP thus breaking that potential routing loop however I want to allow other traffic to be redistributed I need to allow everything else so we need to give a second route map statement for the deny tag 10 route map I'll say route - map deny tag 10 I'll give a higher sequence number this time I'll say permit and we're going to permit everything else no need to match anything because the default is going to match everything now that we've got these route maps created let's apply them to our routing processes let's go into router OSPF process ID 1 I'm going to reissue the redistribute command that I already had for this routing process I'm going to say redistribute ya GRP autonomous system 1 I'm going to include the subnets keyword and I'm also going to apply the route map of Tag 10 what I'm doing is assigning a tag of 10 - routes being redistributed into OSPF from AI GRP now let's apply the other route map to the EIGRP routing process will say router ya GRP autonomous system 1 and let's reissue the redistribute command for the ERP writing process by the way I already have a command underneath this routing process that sets the default metric so I don't need to specify the metric on this command line but I will say redistribute OSPF process ID 1 and going to apply the route map of deny tag 10 you see what I'm doing I'm saying if I'm redistributing it into EIGRP from OSPF I'm going to deny any routes that have a tag of 10 and in this example I did this to one of my redistribution points Rd 1 I won't take the time in this video to show it but to complete the configuration I would need to do the same thing on a router Rd 2 which is my other redistribution point and that's a look at a few different ways that we can troubleshoot a routing loop that might occur when we're doing a mutual route redistribution between a couple of different routing domains and we have more than one redistribution point

Channel: Kevin Wallace Training, LLC

Views: 14,250

Rating: 4.9775281 out of 5

Keywords: Cisco, Cisco IOS (Operating System), 300-101, 300-115, 300-135, ROUTE, SWITCH, TSHOOT, route redistribution, CCNA, Kevin Wallace, Cisco Systems Inc. (Business Operation), 1examamonth, kwtrain, EIGRP, OSPF, BGP, troubleshooting, #kwtrain

Id: -zBFSwvYd0Q

Length: 9min 48sec (588 seconds)

Published: Fri May 15 2015