Cisco CCNP Service Provider 350-501 SPCOR Topic: External BGP
December 21, 2022

1. Basic EBGP Configuration

In this video, we’ll start with the basic BGP configurations. If you remember, we have done most of the IGP configurations in our previous videos. We had some IBGP configurations using full mesh and then also using router factors. And then we have also used peer groups to update those low-back zeros, changing the source, which does a lot of things. So all those things relate to our configuration. So in this section, we’ll move on with some very basic EBGP configurations. So when you come to eBCP configurations, the entire configuration command set remains the same. The thing is, we just need to change the autonomous system number.

So let’s get started here. So I’ll start with router number one. If I verify, everything is already configured with IP addresses as per my default topology. And if I verify, I don’t have any BGP running, and I don’t have any of the other routing protocols running. So the only thing I have is very basic IP addressing as per my diagram. So I’m going to configure EBGP, IBGP, and EPGP. Now, I’ve got a diagram here. You can see in this diagram that I have three routers, and I’m planning to configure these two routers. It will be a part of autonomous system number 500, and then this router will be in ASA 600.

 So we don’t have any router vectors for those things. So we just have some plain configurations. We must now configure router number two. Router two is the router that is forming an EBP neighborship with router three because the latter is in a different area. And also, it is forming an IBGP neighbourhood with Router One because they are in the same area. And because the router one only needs to establish a neighbourly relationship with the routers within the same year, the router one can only form a neighbourly relationship with the router two. And from router one to three, we don’t have any EVP required. So we are not going to do that. Then, router 3 establishes only one EBGP session. So let’s see the configurations here.

So I’ll start with router BGP. Router One: I’m going to run Router One in the 500. I’m going to say no order summary, no synchronization. And then I’ll say “neighbour one,” which is my router two. And then I’m going to say 500. So my autonomous system number and my neighbor’s autonomous system number are the same. Now the router will understand that it is my internal BGP neighbor. After that, I’m going to use my Teredo network, which is my LAN interface. In addition, I’m rosing my single interface. So I only needed two interfaces as per my diagram. The one-dot network and ten-dot network with the default subnet mass are done in a similar way. If I go to router two on the router, I have to configure two neighbour commands, and the router is running autonomous system 500. No auto summary, no synchronization, and then I’m going to say “neighbour one” and “remote” as 500.

Then there’s neighbour 2, who’s my router 3. Ace is 600. Now, you can see here that my autonomous system number is 500, and I’m going to pair with router 1, which is also 500. So the routers will treat this as nothing, but it’s an IBGP neighborship. And when it comes to router 3, it’s going to treat it as an EBGP neighbour because of the different autonomous system numbers. And then I’m going to advertise my LAN interface, the router’s two LAN interfaces, and my VAN interfaces. Done. So on the router, I just configured two neighbors, which is very basic. Then I’ll request my LAN interface as well as two VLAN interfaces on the router 3. Also, I’m going to do the same thing. Router three is running autonomous system number 600. Then I’ll say no autosummary, no synchronisation neighbour two, which is my router two remote, as well as 500. And then I’m going to die along with my LAN interface, a 30-dot network, and then one interface. Done. So now that we’ve finished configuring these basic neighbours and network commands, the first thing we need to verify is the neighbour ship. So we will go and check on router two, because router two should have two neighbors. I should see neighborship from both ends if I show the IPBG summary. Router one, the neighbourhood is already up on the other end. The neighbour relationship between routers three and four is now active as well. Now, just now, as you can see here, from each neighbor, I’m going to receive two routes. And if I verify the shy PGP, I can see all the routes.

So in this output, I’ll get into more detail when we come to the path selection process. And if I verify on router three, if I get Shuip route BGP or I can see the routes coming from router two as well as from router one, And this is the connected interface between router one and router two. So maybe we get the routes and we see the routing table here; you can see 20 because all these routes on router 3 are coming from router 2, which is my external BGP neighbor. So which means any routes coming from external BGPneighbor, it has the default administrative distance of 20, which is more preferred than if you are learning the routes from the same OSPF or EHRP. Also, this is more preferred. Now, if I check on router 2, if I can show IP route BGP on router 2, you can see a 10-dot network coming from router 1, which is the same as the default administrator distance of 200. And the 30 dot network is coming from router three, which has an administrator distance of 20.

So because this route is learned from external BGP neighbors, the default administrator distance for any EBGP route will be 20. And any route learned from the internal BGP neighbourhood will have the administrator distance of 200. Because internally, within the same year, we always want OSPF EHR rip protocols, which are IGP protocols, we want those protocols to be preferred. So that’s the reason internally: BGP is not and will not be used for the selection of the best stud. But BGP is specially designed for extreme gateway protocols. And you can see it’s two two at the next stop. If I go to router one and verify the routes, I’m able to see the routes. But you can see the three-dot network. The next stop is not router two. That is one dot or two. It has two dots. Two dot two. So, regarding this, I’ll explain it to you in my next video. We’ll see more detail in our next video because there is a default rule in the BGP’s next top we’ll see more in detail in our next video. So, from this lab, we can understand very basic EBGP configurations.And if you see the configurations of a router, if I verify the configurations of any router, say on the router, I’m going to verify the configurations look similar to your standard IBGP configurations, with the exception of changing the AS number. That’s it.

2. BGP Using NextHop Self

In this video, we’ll try to understand the default PGP Next Top behavior. So next stop is one of the attributes that can be used in the best path calculation, and it is a well-known and managed attribute. So whenever a BGP carries its routes, it is going to carry the next stop information. But the behaviour of the next stop is slightly different when compared with our normal routing protocol. I got some points here, for example. BGP is a biassed protocol, not a router-by-router protocol. As a result, BGP will only change the next-stop when the autonomous system number changes.

Okay? So let’s try to understand the difference here. I’ve got a very simple diagram here. You can see that I have an autonomous system 500, 600, and four routers in that. And I just named those routers B-C-A-B-C-D-E-F-G. Some names. It can be any IP address. I’m not listing any specific IP. Just to make it very simple, I’m going to name them with some of their respective alphabets. Now, what I want is to understand the default routing behaviour and the default BGP behavior. So we’ll compare, and then we’ll try to figure out what the BGP’s default next-stop behaviour is. So take an example: I’ve got a route here, the 10-dot network. Now I want to ensure that whenever BGP is inside the BGP, router A advances to router B because it is an external BGP neighbor. Depending on how you configure your IBGP neighborship here, router B will address Ca as well as raise Das and go to E. Because B will send to everyone if you use full mesh. If you use router factors, it will send to our server and then to the destination. But anyway, the update will go to all the neighbors, and from the router E again, it will go to F. And F is going to forward this to G. Now that it is forwarding, it is also going to carry the next-stop information.

like the network will be anywhere. It will be advised along with the next-stop information. because BGP also carries the next-stop information. So, what can we learn about normal routing? So this is our normal routing. So a 10-dot network will be advertised from router A to router B. Now router A is going to advance to router B when it is advertising the 10-dot network. Who will be the next operator? Because router B receives it from router A, the next operators will be routed A. So that means the next stop orders will be at this address, whatever the IPRs. Now B is going to forward it to router C. Now C is going to receive the same network, but what is the next-stop address? The top order will usually be B, and C will pass to D. Let’s say in the case of normal routing, the next stop address may be C, and the next operator may be E. Sorry, D. So, if it is going to router D or E, or if B router is sending to E correctly due to full mesh for all routers, the next stop orders will be B only. Right? Or it can be C, or it can be T. So I’m just taking it as a normal route. So in case of normal routing if you remove the BGP config BGP part the next operators will be B here and on the router D.

The next stop order will be C, and on route E, the next operator will be T because De is going to receive the route from D, which means the next stop order is going to be D. From there, the update goes to F. F will write the next operator as E, and G is going to write the next operator as F because next stop is simply next router. Right. That is what we learn from our normal routing configurations. If I use any of the routing protocols, like OSPF, EHRP, R, or IP, the next stopbuttons will always be the next router letters. That is something called normal routing. So we are talking about normal routing that we learn. So now we’ll try to understand the behaviour of the BGP. Now. Now we are going to verify BGP behavior. In the case of BGP behavior, how will the next stop be determined? Now, when the router B receives the next stop from us, it will be A. When router C receives the next stop, butter will be A once more. Now try it again. Now the next stop order will be E and the next top order will be S, so now we need to understand what this exactly means. In the case of BGP, there is a rule that states that whenever an update is sent from one IBGP neighbour to another, the next-stop orders will not change.

The next stop order remains the same. When you update, you are sending from one EBGP neighbour to another. This means that when it goes outside the Is, it will change the next top behavior. In the case of BGP, the default behaviour of the next stop is to change the next top address only when the autonomous system number changes. However, it will not change the next stop when advertising to the same routes within the same autonomous system. So that is something we need to understand with this scenario here. So let me show you again here. In my diagram, the next stop is at us. Now this is route is coming from external BGP neighbor. So when a route is coming from an external BGP neighbor, it is going to write this as the next stop. That is again the same in both cases—normal routing as well as BGP routing. However, whenever an update from B is sent to C, D, or E, the update is sent to all three. which means whenever it is advertising the update, it is also going to advertise the tendon network.

At the same time, it will not change the next top order because when BGP is going to advertise any network, it will also include the next stop information. But when it is advertising to the same autonomous system numbersor the routers within the same as it is not going to change the next stop which means the tendon network will be addressed to router C with the same next stop Aand then same next stop A, same next top A.When the same update for the same Tendon network is sent to another EBGP neighbor, it will enable EBGP. Now, when F receives, F will receive the next stop address as the next top order, which means that now here it changes, but when the router sends to Gagain, the next stop address will not change. Now this is something that is the default behaviour in BGP. BGP will only change the next stop when it changes the autonomous system number. Now B is going to receive the update, and he is going to forward the same next stop to all the routers inside S. Now, this might affect your network’s reachability in some cases because, for example, I got the same diagram here also. Now let’s take an example. This network is, let’s say, 15 to 1. This is the IP address of this network. Now router C is going to receive the network with a 10-dot network, and the next stop address is A—nothing but 15 one one. Now if the router C does not know how to reach the next stop, then it might affect your reachability, so you will receive the route, but you won’t have the best route, and you will not be able to communicate with the TENT network, which is in a different case. So there are now two possible solutions to this type of problem.

So the first solution is to advertise all your van interfaces, which is very little because advertising the via interfaces again means you need to advertise all these van interfaces inside the BGP so that router knows where that 15-1 address is so that it can forward the packet directly to that address. Or what I can do is change this next-stop behavior. Now we can manually change the next stop. So to manually change the next stop, we need to go to the border routers. Here. I need to go to router B, and inside the router BGP500 or 600, we need to go to router BGP 600.And then I have to send neighbor, whoever that is. In my example, it is C, so it can be an IP address. Actually, neighbour C, I’m going to send the next hop command. So when I give this command automatically, I’m saying that whatever updates I’m sending to this neighbor, please change the next stop address to my address. As a result, the next stop address will be A by default. Now it will change to B. Similarly, you can do the same thing for all the internal labels that have neighbours pointing to Dand, and then I’ll say next option. Similarly, I can say neighbor, pointing to E, and then observe. So, by issuing this command to the border router pointing to our internal routers, we are manually altering the next stop behaviour of these devices. That’s the reason we use the next-stop self-command. So the next-stop command is really required if you want to change the default next-stop behavior.

And the default next stop behaviour is that, by default, within the IS, the next stop will not change. So what I can do is say “next stop, myself” on this router. So the next stop order will become B on this router. Now if the next stop at us is B, then probably the E router knows how to reach B because internally we are also running an IGP, and in that IGP, we are advertising those networks. Maybe the OSPF I’m running is just me advertising my networks inside the OSPF. He knows where B is now that he is outside. So automatically, it will not affect your reachable team. So it’s something we need to keep in mind. If you see the route in the routing table, if the next stop is not reachable, you’ll not see the route in the routing table, but you’ll definitely see the route in the BGB table. So definitely, it will affect your reachability. There are two possible solutions. Either you advertise the interface or change the default next stop behaviour by using a command called next stop. Sir, this is something that is recommended. Now, similarly, if you are doing it from the other end, let’s say you have a 50-dot network here. Now that you’re receiving the route from G to F and from F to E, you’re getting the route and router that E will send to D without changing the next stop. But I can go to router E, and I can say, “Neighbor D, next stop, myself.” I can do that. Similarly, I can do the same thing on the other end as well.

So this is the default behaviour of the next stop. So it’s really important to understand. Now we’ll try to verify the same thing via lab. Now, if you remember from our previous video, we have done a very basic EBGP configuration for this diagram. So I’m going to keep doing that, and I’ve already done these EBGP configurations. Now I’m going to verify the default next-stop behavior. So let’s go to my console screen here. So I’ve got my diagram here, and this is my screen. And if you verify our previous video, we did these basic EBGP configurations where router one is already establishing neighborship with router two and advertising the van interfaces. Now, anyway, I’m advertising the one interface here. As a result, I believe it will have no effect on reachability. But we want to verify the default next-top behaviour here. You’ll also find two BGP configurations if you verify my router. Now router two is already configured with neighbour one, one in 500, and neighbour three, two in 600. And on the router, this is pre configured.So if you want, you can continue with the same lab as what you did over there, or else you can start from scratch and make sure that you finish this configuration. So now if I verify on my router too, I can see both of my neighbors, and from both of the neighbors, I’m receiving the route. So now we’ll verify the next top behavior. Now, I’ll say “Show IP BGP,” or “Show IP Route BGP.” If you try to see here, the next stop address for the 30 dot network on the router one is two two. Now, by default, So here’s a very simple scenario here. So this 30 dot network is advised to router two, and router two receives this 30 dot network, and the default next top address will be: what is the next top address?

Two dots because this network is resilient. That is the configuration of my router’s three connected interfaces. Now what’s happening here is that router 2 is sending our bit to internal BGP, enabling the same network but without changing the next stop. And the next stop address is “2” only, without changing the next stop. So that is something that is happening here. Now, you can try to verify here or use the show IV BGP BGP table here. Also, you can see the 30 dot network I’m getting, and the next stop address is two dot two dot, two two on the router one. That’s my next stop. Or you can verify with this command more in detail. You can see that the next top order is two two. It’s coming from one to two. Okay, so this BGP table and those things, I’ll go over in more detail when I start my path manipulation process. Now in this case, I really don’t have any problem communicating with 31 One, which I can still access because I advertise the one interface, if you remember. If I check my routing table, router one has to go through two dots, two dots, and two dots to reach this network, and it knows where the two-door network is. It is on this via one dot one do one or two. So now, in this case,

 I don’t have any problem with communication because I saw the advertisements for van interfaces here. Let’s say I don’t do that. So what I’ll do is try to go to my router 3, and I will remove that two-door network from my advertisements. I’m going to do the same thing with router number two. On Router 2, I’m going to remove that two-door network from the advertisements. Now, if I verify, you can see this two-door network is not present. Even if I confirm the three-door network here, the next stop for us is this. And if I verify my routing table, there is no three-dot network because to reach this three-dot network, the next stop address is two, and router one doesn’t know where two is because of this default next stop behavior. Now, advertising all the interfaces is really not possible. So either you can go with that option, but what I want to do is change this next-stop behaviour to our normal routing behavior. So I’ll go to router number two, then to router number two again, and say, “Neighbor, what’s your name?” My neighbor’s neighbour is one, and I’m going to say next stop, self-command. That’s something I’m going to do.

So I’ll go to my router two, and then I’ll say router BGP 500 and point to my internal label. So I’ll finish up the border routers that point to my internal routers and then say, “Next stop, self command.” So now, when I verify my router, it will use shy BGP or shy PGP. So previously, as you can see, the next top address was this. before I configured the next stop command. Now if you verify the next stop address, it automatically changes to router two because I have verified it manually by saying the next stop command. So now if I verify my routing table, so IPGP, you can see this three-door network route; you can see the route in my routing table, and if I try to ping, I should be able to ping. So this is something very useful. This is a very basic lab for understanding the next term’s behavior. When you’re working with large networks, you definitely need to consider these things. So, whenever you receive a route from an external BGP neighbour and then advise it to internal BGP enablers, make sure that your next-stop address is reachable in case the next-stop address is not reachable or this man interface is required to rise. So make sure that you change the next stop address to your local router address within the same range by using the next stop subcommand.

Leave a Reply

How It Works

Step 1. Choose Exam
on ExamLabs
Download IT Exams Questions & Answers
Step 2. Open Exam with
Avanset Exam Simulator
Press here to download VCE Exam Simulator that simulates real exam environment
Step 3. Study
& Pass
IT Exams Anywhere, Anytime!