1. BGP Introduction
So welcome to BGP video trainings. Now, in this section, we’ll start with some of the basics of PGP, like basic concepts and terminology. So, if you just want to look at the basics, we’ll start with the autonomous system number. What is an autonomous system number? And then we’ll see the differences between IGP and EGP protocols and some of the basic features of BGP. And then we’ll see the loop prevention mechanism in the BGP. And then finally, we’ll wrap up this video by understanding when it is more appropriate to use BGP and when it is not really recommended to use BGP. So let’s get started. first with the autonomous system number. Now, from our basic CCN studies, we have already learned what an autonomous system number is Autonomous System Number.
The autonomous system number is simple. It’s a collection of networks under a single common administrative domain. So that’s what we say. It’s a collection of routers under a single administrative domain. As an example, I go to an organisation where it has specific locations that are all part of the same organization. Let’s say it is a number 100, and this organisation is, let’s say, ABC, and all the routers within that organisation will be represented by one number. And we call that an autonomous system er. And we cow, the next thing is similar to the way we can have multiple autonomous system numbers, and if you want to communicate within the same autonomous system number or within the same organization, generally we use IGP protocols. Like whatever the protocols we learn in ourprevious videos, in the basic CCNA CCNP videos. As a result, we refer to Rip. Rip version two (iGRP is no longer used). EHRP? ISS’s OSP These are all IGP protocols. So the main difference is that IGB protocols operate within the same autonomous system number.
So if you want to communicate within the same organisation or within the same autonomous systems, then you use something called interior gateway protocols. We call them IGP protocols. So all these protocols listed here are typically your IGP protocols. But let’s say you want to communicate between two or more different autonomous system numbers, then we use an exterior gateway routing protocol. And the only protocol that is running on the Internet that allows you to exchange routes between two or more different autonomous system members is the BGP protocol. So BGP is the only protocol running on the Internet backbone and is primarily responsible for maintaining the routes or exchanging the routes between two or more autonomous systems. And the service providers use BGB protocol and busing BGP, they control all the routing information. So BGP is the only protocol that is designed for the huge network that is the Internet. So we’ll see some of the points anyway, more into that, more in detail about the BGP features.
Now, in this section, we have seen the Autonomous System number. It suggests routers are under a common administration. Now, if you are communicating between two or more different autonomous systems, like we are trying to see an exchange route between two or more different Adam system numbers. We use a protocol called PGP, okay? So the next thing we’ll see is that we’ll get into IGP routing. Like when we talk about IGP routing, we have seen static default and dynamic routings. And we’ve seen Rap Ehrposp of ISS, all these protocols, inside that dynamic routing. As a result, all of these protocols operate under the same autonomous system numbers. So we call them IGP protocols. But if you want to communicate between two or more different autonomous system numbers, we use the BGP protocol. So first, we’ll start with some of the basic features of BGP. Like BGP, this is a standard protocol. You can run this protocol on any vendor device, and it is an exterior gateway routing protocol. exterior in the sense that it is going to exchange routes between two or more different autonomous system members. So we call it the “exterior gateway protocol,” and it is specially designed for inter-domain routing. “Inters” means communication between two or more different autonomous system members. Assume this is 100 and you want to communicate with 200 people. As shown in the previous diagram, we use the BGP protocol, which is specifically designed to scale a large network like the Internet. It supports class lists, and then it supports all your FLSM VLMs here, the Dr. Manual, and auto summarizations. All these things are supported in the BGP, just like they are in all our IGP protocols. And it’s similar to your internal protocol updates, where incremental and trigger updates are supported.
And we call BGP a path-vector protocol. What exactly does “prod vector” mean? So, a path vector is a method of extending routes along with path information. Consider the following example: I have an adiagram here; you can see the diagram here. Now let’s take an example. This router will publicise this zero-dot, zero-dot network. So now this router is going to advertise this network into the next autonomous system, which means it is going from one as to another as. So it is going to carry that ten-point, one-point dot network, and at the same time, it is going to carry the path information. So it is coming from, which is 65,400 as Now when it sends, it is going to pass on this information to this router. And when this router is going to pass on the information to the next router or the next autonomous system, it is going to carry that network information just like your normal IGB protocol, but it is going to carry the autonomous system path information. So it’s going to say that it’s coming from 65,400, which is this, and then it is coming from these two S. So it’s crossing these two autonomous systems and reaching here. And finally, when this router is going to send it to another router, it is going to pass on that information like it is originating from this, then it has reached the next, and finally it has reached this, which is the last s. So we call this “path information,” and when it is going to carry any routing update, it’s going to make sure that it is going to carry this autonomous system path information.
This is extremely useful, especially if you’re in a loop prevention mechanism. There is something called a “loop prevention mechanism,” which I’ll be discussing in our next video, in the next sections. I’ll probably get into that more in detail. Okay, so we call it “path vector protocol,” where it is going to carry the path information from which autonomous system hops it is moving on. So that’s the reason we call it “path vector protocol.” And then there were some more differences and some more features of BGP. It uses unicast to send updates to manually selected neighbors. So at this point, I’ll get more clarity when I get into BGP neighbors. Now this point will tell you that, like manually, you have to convert the neighborhood. Now, what is “manually”?Let’s take an example. We have two routers, and if I’m running OSP or another protocol, say EHRP, by default, when I advertise this interface, it will send a hello message to the other side of the interface, and then it will respond to that hello message. Based on those “hello” messages, they automatically establish a neighbouring relationship. So they’ll automatically become neighbors, and they’ll automatically build a neighbour table. There is something like automatic neighborship, but in BGP we don’t have that. In BGP, we have to manually configure the neighbor. which means we need to say that router two, you are man enough. Let’s say this router one is the neighbour of router two.
On router one, you have to configure a neighbour command saying that you are maneuvering, and on router two, you have to configure a neighbour command saying that router one is the neighbour of router two. And if the neighbour commands on both sites match, and everything is properly configured, the neighborship will appear; this is what manual neighborship configurations are. We established some initial labs that were entirely dedicated to this. So I’m not getting into practical things. Probably in our next couple of videos, in the third or fourth video, we will directly jump to this lab. where I will also show you the configuration, how it works, and some more detail on whatever we discussed. So next thing, it’s a BGP application-level protocol for reliability. It operates on port 179 in the metric system and employs the TCP protocol. It uses a lot of attributes. So we have a weight attribute, local preference, and path origin. Next up, we’ve got a lot of attributes. BGP supports very rich emotional attributes, which can affect the path manipulation process. Like in OSPF, it uses bandwidth, and the default formula it uses is ten to the power of eight divided by bandwidth. In the case of EHRP, it is going to use bandwidth delay, load, and MTU reliability.
Hopkins will be used in the case of Rap. But in BGP, the pass selection process varies and relies heavily on these attributes. So, more on these attributes. We’ll definitely get into this in our advanced BGP concepts. So as of now, I can simply say that it supports rich attributes, which can really affect the path manipulations. And then finally, the administrative distance is 20 if the route is coming from external BGP; again, we’ll see the differences between external BGP and external BGP. Also, if the route is coming from external BGP or from a different autonomous system name, that will have an administrative distance of 20. If the route is coming from internal BGP, we have an administrative distance of 200. So again, more detail about IBGP and BGP. I’ll definitely go into more detail in our following sections. So here I’m just getting into some basic information that is going to define the features of BGP. So, as you may recall, I recently discussed loop prevention mechanisms in BGP.
There’s a chance you’re connecting to the Internet and connecting to your service folder from one side while also connecting to another service folder from the other. So you might be advertising your route; come to our network. And this route is advised to a service forwarder or to some other autonomous system numbers, and from there it reaches the Internet Cloud, with the possibility of the same route returning. Now, that way, I can create the loop. However, in BGP, there is a rule that whenever a router sees its own number, such as when I discussed path vector behaviour earlier. Take, for example, this network (1810 plus 16). This network is now starting with the number 100. When it sends to the A router the second AIS, it is going to carry the network information, like 180 or 10, whatever it is, and the autonomous system path information, saying that it is coming from 100. And from there, when it reaches the other Now the same network: 180, 1016. It’s going to carry the as Path information that it is coming from 100 here and then reaching200 and then finally reaching to my S.When this occurs, it will be forwarded to ZIP code 500. It will forward that network information with 180 and state that it is coming from 100.
So in general, it will be reverse and then it’s reaching a S 200 here and then coming to 300.So these routers will now understand that there are 500 routers inside this network; they will understand that this network starts with 100, then 200, then 300, and finally reaches my autonomous system members. This is now advertised as 100 from here onwards. Now, when it advertises this information, when a router receives this information, it is already running as route 100, and whenever it sees its own as route number within that update, it will reject the route. So this is the default loop prevention mechanism in the BGP protocol. When a router behaves as shown in Figure 180, the tendot network is not accepted by the 100 prefix or in its as path. In the BGP, this is a loop detection and default loop prevention mechanism. So BGP will never instal any route when it sees its own as a member. So I’m running as 100 and I’m getting a route called the “Ten Dot Network,” which also has 100. So it’s something originating in MYS and coming back to MYS again. So, because when you are connecting to the internet, you are definitely connecting to multiple routes, and there is a possibility that the routes will be coming from multiple sites or that the same route is coming back again to MYS. So this is some kind of loop prevention mechanism. You must understand the BGP’s default behavior.
2. When BGP is More Appropriate
Now, finally, the last thing we’ll see here now that we have seen some of the basic features First thing, we started with a number and then IGP routing BGP features, and this is what we discussed just now: the path vector behaviour and then the loop prevention mechanism here. Now the last thing we’ll see in this video is when BGP is more appropriate to use. You may now find yourself in a position where you are in charge of an organization, and your organisation is most likely using BGP. So if you’re running a big organisation, you might be running BGP. Large organisations typically use BGP for path manipulations in order to exchange out of service for them. Now, first, we need to understand when it is more appropriate to use BGP. So, if I decide to use BGP, I must first determine whether I can use it, cannot use it, or whether it is required. So the first condition for using BGP is that you should be a service folder. So if I’m the service provider, let’s say I’m working for any service for the network, then servicefour is going to provide services to many customers, which means he is going to connect many customers to different members of the autonomous system members.
So to exchange the information between these multiple autonomous system numbers, it becomes mandatory for the service provider to run BGP because BGP is the only protocol designed to exchange the route between the multiple layers and it can do some path manipulation. So it is specially designed for that. So the entire internet So, whatever the network, we can say that BGP is used by the world’s largest network, the Internet. As a result, service folders keep the Internet running. So service folders connect to each other, and they maintain the biggest network. Now, if you want to access the Internet, let’s say I’m going to [email protected]. When my request reaches my ISP, and from the ISP, we say it is going to the Internet, So the Internet is a public network where everyone is connected.
So that means this ISP is going to maintain where your Yahoo server is, and it’s going to maintain the route of the Yahoo server. And the Yahoo server is located elsewhere—possibly with a different company or organization. So it’s going to send a request to the Yahoo server, get the reply, and finally send it to ours. That is, if you are a service provider, you must maintain the routes of various autonomous systems, or everything necessary to maintain the routes. So this can be done by using the BGP protocol, because in this scenario we are not going to use any of the IGP protocols because IGP protocols are only designed within the same autonomous system number. But when you talk about the Internet, we are moving between different servers to reach the Yahoo server and coming back. So maybe I’m just moving around multiple addresses—probably more than 56—to reach that particular Yahoo server.
So that’s the first case. If you run an ISP, or if you are an ISP, you must use the BGP protocol. Okay? So in case I’m not an ISP, let’s say my company decides to run BGP even though it’s not an ISP. Now it is really recommended if you have multiple autonomous systems and you are connecting to multipliers, which we call a “multi-homing environment” and where you need to do some path manipulations. So let me take one simple scenario to understand this second point: when is BGP more appropriate to use, okay?So I’ve got a diagram here; I’m going to keep on diagramming. Let’s say this is my organization, which is connected to I got many routers, plenty of routers here, okay? So now your organisation is simply connecting to your ISP, and we are connecting to the Internet from there. So we just got only one connection to ISP.
In that case, it’s really not recommended to use PCP because normally what we do to access the Internet is use the default route. So towards the ISP and from the ISP, we confer a static route, and we redistribute that distorted route into your IGP, and you will simply exchange the route without any problem. This is one common scenario where we use it. This is the most common scenario without BGP. So especially in small or medium-sized networks, we generally follow this scenario where we have some IGP protocol running, probably EHRP, OSPF, or any of the other protocols. And we have a default route pointing to our ISP. And then we are redistributing the static route into our IGB protocols, and we are just allowing all the routers in MYS to access the Internet via the same ISP. So we call this a single-home environment, where you have a single exit path towards the service or order, and we just have one exit path. As a result, we do not recommend using PGP in this case. But let’s say this is my organization. It’s a very big organization. In my organization, I have a large number of autonomous routers. I also have several exit options. I’m most likely connected to two service portals, ISP one and ISP two. So maybe for redundancy, I have gone with multiple service providers. From there, I’m going to access my Internet, or probably the resources of different organizations. This is my ABC organization.
So we call this a multi-homing environment. As a result, you now have multiple exit routes from two different directions. You can now try the same thing here, such as just a default route or a primary default route backup. And you can use a normal IGP to reach any one of these. But what I can do in this type of scenario is use BGP to do some path manipulations. What I can do is say that some of the traffic should use this route. Assume I have a 10-dot network, a 20-dot network, a 30-dot network, and a 40-dot network in my production network. So I can manipulate the path so that this tendot network and 20-dot network use this route to go outside my autonomous system numbers, and then I can tell the remaining two networks to go from this route. So we call this path manipulation. Now this is something BGP is going to provide you with. So if you want to do something like this, then you can use BGP and instal the route. So now we are not going to use a default route here. We are just trying to exchange the routes in the service folder. And, despite the fact that I’m also running BGP and the service board is running BGP, I’m going to instal all of the routes in the service folder, or routes that come from the service folder, into my BGP routing table. As a result, some of the larger organizations, in particular, adopt these practices. We call this a “multihoming environment,” because you have multiple exit paths and we’re going to manipulate them. So this is the next case where we can use PGP. So, in greater depth I’m getting into that, like the different types of connections to an ISP and the different types of configurations. I’m getting into detail about that in my next video. So here I’m just trying to figure out when it is more appropriate to use BGP. So if you are a service porter, it becomes mandatory for you to use BGP.
And if you’re not a service portal, if you’re running a business in a multi-homing environment, and you want to manipulate the path of traffic entering or leaving your au, So in this type of scenario, I can use BGP, and it is not recommended to use BGP if you’re running a single-homing environment. If you recall, as I mentioned in another example, MYS is only connecting to one connection to one of the ISPs, and there is only one exit path. So, in this case, BGP is not required because you simply configure a default route to the service for order because you already have one route. So we don’t really need to do any path manipulation. So it’s really not recommended to use BGP in this type of scenario. And if you lack resources, Because if you are running BGP in your production networkmeans you are going to run the BGP and youare going to instal all the routes of internet probablycoming from service portal, which means you should have avery high end devices, probably your core routers, to maintain such a big routing table and to process it.
And also, if you have a limited understanding of PGP or if you’re really not sure how the BGP behaves, how it selects the best path, or how it’s going to work, troubleshooting BGP So it’s really not recommended to use BGP. So that is something that is recommended to keep in mind when it is more appropriate and when it is not appropriate to use BGP. So finally, we have seen some of the basic concepts of BGP here. So initially, we started with the autonomous system number. What is an autonomous system number? It’s a number that identifies a company. Then we saw the fundamental differences between high-GP and BGP, as well as some of the fundamental features later on. And we have seen the blue prevention mechanism. The default loop prevention mechanism in the BGP is wheneverany route is getting into the as, it will check if it sees its own as number inside that it’s not going to instal that route. And then finally, we have seen when to use and when not to use BP. So probably in our next video, we’ll see some more and get into some more detail about the BGP configurations.
3. BGP Options – Connecting to Internet
Now we are ready to get into some more details about the BGP here. So, if you remember from our previous video, we learned some of the basics of BGP, where we started with autonomous system numbers, IPGP, and BGP differences. And then BGP features a loop prevention mechanism and explains when to use BGP and when not to use BGP. So this is something that we have seen in a number of first videos. Now, we are going to continue from there. Now, in this video, we are going to learn about the different types of ISP connections.
So depending on the size of the company and the company’s requirements, we can have different types of connections to the ISP. And we have something called single-home, dual-home, multi-home, and dual-multi-home connections. We’ll try to see the differences between these connections and determine which type of connection, which type of routing, or which type of BGP configuration is more appropriate. We’ll try to look into that, and then finally, we’ll see if you’re connecting to your ISP. Let’s say I’m connecting to my ISP via a BGP route. What are the different methods we try to connect? Okay, that’s what it is again; we’ll get into more detail about it later. For example, we have three options: default routes, specific routes, and exchanging all routes. So we’ll see these two major themes in this video of training. So let’s start with the different types of ISP connections. So the first one is a single home connection. Now, if you remember from the previous video, I already discussed single homes. So in a single-family home environment, we have just one exit path. So this is my autonomous system number—let’s say autonomous system number 80. It’s connecting to my service folder, which is my ISP one.
And I only have one connection. So in this type of connection, we can just have a single default route. I can connect to this router. I can simply say “IP route zero zero” and whatever. The next stop is, let’s say, one or two. And then from the ISP side, he’s going to configure a static route for the public IP, whatever we are using here. So it’s just a normal default configuration, like what we did in our basic CCN. So in this section, in this single home environment, it’s really not recommended to run BGP. We just have a site with a single ISP connection and this is fine.
That is not heavily reliant on the Internet or a broadband connection. So we can use static routes, which are the most commonly used, or even some of the default routes provided by the ISP. The second connection is something called a “dual homing environment.” This is now more common in the majority of companies that do not use BGP. We generally have two links connecting to service portals. What is the benefit of connecting to two links? The main benefit is that if one of the links fails, we can go to this router and enter IP route zero here, as well as zero, zero, zero. I’ll say next stop addresses 1, 1, and 2, which is my first link, and I’m going to confirm one more default route. On the same router, I’m going to say IP, and then I’m going to say two with some different administrator distance. Now, by default, it will use the first link because of administrative distance, default is one. If that primary link fails, it is going to use the second link. Now, the main advantage we get from this dual home site connection is redundancy.
The major advantage We can now do this in cases where you are connecting to two different routers or have a connection to two different routers for redundancy. And we generally run some HSRP, VR, and RPG to have multiple gateways again in the land, so that’s something different in terms of configurations, and then that’s how we do it in the case of dual homing. And in this scenario also, it’s not really recommended to use PGP. So we’re not going to run any BGP because we’re just connecting to the same service for orders. As a result, if any of the links fails, only the backup link is used, and we have very little path manipulation here. As you can see, we have very limited path manipulation options. So in this scenario also, it’s not recommended to use PGP, but when it comes to multihoming environments, like in this scenario, we can use BGP here. Now, in a multi-homing environment, with two exit paths from the same as this is my S, it’s connecting to ISP one and connecting to ISP two, and I can tell some of the traffic.
Let’s say I have four networks, such as a 10-dot network, a 20-dot network, a 30-dot network, and a 40-dot network, and I know that the 10-dot network should use a primary ISP and the other two, which are my 3-dot networks, can use an alternate secondary ISP. If this link fails, everyone is routed through the secondary link. So this is what we can do: some path manipulation where we can decide how the traffic should enter the IS or exit the S. So we call this path manipulation. So in this type of scenario, BGP is more typically used, and we generally exchange the routes. In this scenario, we’re not simply getting the default route; we are not sending only the default route here. Instead, we are also exchanging all the routes from the service folder through the PGP protocol. But in the case of the previous two networks, we are just sending the default route. Any unknown packet will be sent to the ISP, and the ISP will return that packet based on the static route. But here we are trying to exchange all the routes through BGP or maybe some specific routes and then we have something called dual home environment.
We just got some extra links that will provide more redundancy—the most redundancy. And, once again, we can use BGP to perform path manipulations. So these are the four different types of implementations we can use, and especially in the first two networks, the first two types BGP is not really recommended because in the first two scenarios, what we do is simply configure a default route, like default routing, if you remember the basic default routing in the basic CCN studies. And from the ISP side, we get a static route for the public IP because our router also does some NAT translations. That’s how it works.
But whereas in the case of the remaining two setups, we are not only connecting to ISP one for redundancy, we are also going to connect to ISP two also. So I can tell you that some networks should go from here and the rest should go from there. If you want to do that path manipulation within your application, we need to run BGP. Anyway, the service folder is going to run BGP. Now you need to have a registered number, and based on that number, I can exchange the routes between them from the service folder, and I can do some path manipulation where I will decide how my routes or how my network should go to the internet via this route or that route. So this is something we get if you are using BGP, especially in the big networks, where they use BGP for these things. Or if you are working for a service folder, you may have some customers with this type of connection and need to do some path manipulation. That okay.
For customer 1, they should exit from one path, and for customer 2, they should use the alternate path. So that’s something really useful. BGP is designed for that only. So the next thing we need to figure out is how we’re going to connect to the service and exchange routes. Now there are three different methods of exchanging routes, communicating with the service forwarder or routes, or communicating via the internet. So one option is to just have a single default route, and the next option is to have some more specific routes plus the default route. And the third option is to have the complete routing table. Now, how it is going to work. So I’ve got some diagrams here. So these diagrams look a little bit complicated. So I’ll try to make it simple here by using my own diagrams, as you can see. So what I’m going to do is use this as my autonomous system number, and I’ve got plenty of routers, probably more than 50 or 60 routers connected in my S, and my S number is 65,060 5000 subneumber. And I’m going to connect to two different ISPs.
Already, I have taken some tools from ISP One and ISP Two. Okay? So one option is what I can do if I want to communicate with my internet or with anyone else outside. Now one option I can simply do isI can simply use a single default route. That is something I can do; all I have to do is go to my router and confirm a default route here pointing to my ISP. And then, from the ISP, he will confirm a static route. And then I’ll redistribute that default route into my IGP. Now, based on the IGP, the same thing I’ll do here Also, I’ll point out a default here, and we have a static route from the ISP that I’ll redistribute into BGP. And now I’m getting the default route from both sides. I’m getting the default route from both sides. So based on the IGP metric, like the OSP of EHRP, it is going to use any one default route as my primary route. If that route goes down, then automatically it is going to use the second route. That is the first option. This option has an advantage, but it also has a disadvantage. Now the advantage of this first type of implementation is that we don’t really need to know the routes. Let’s say I’m going to try to access a Yahoo! or Gmail server on the Internet. I have no reason to keep any information about that Yahoo! or Gmail account.
So all I’m going to keep is a default route. So any unknown packet will be sent to the ISP, and the ISP will take it and forward it to Yahoo. And your communication will begin between them. That’s a very good advantage. With that, if you don’t have enough resources, you can really try this. But the drawback is that we cannot do any path manipulations. So path manipulation is not possible in this scenario. Because even though you are going to use two default routes, it’s going to use any one of the route as primary route based on IGP metric. And if that primary route fails, then only the alternate route will be used. That is not possible. That is the main disadvantage here. Now it depends upon whether it’s a very good option. Especially if you don’t have enough resources in your network, you don’t have high-speed routers to maintain the routes to run BGP. You can simply continue with this option. This is extremely beneficial, particularly for small and medium-sized businesses.
They go with this kind of solution. That is the first option we can use as a simple default route. The second option is that we can have a default route; that is one thing. We can also have some specific routes. This is an additional method of implementing this one or more implementations, which we can do if you’re using BGP. Now, what I’m going to do is have a default route from this side, and from both sides, I’m going to have a default route that is just like a case. At the same time, I am also receiving some specific routes. Specific routes imply that, for example, I frequently access some of my servers on the Internet. There are some ten servers, or my company’s servers, or anything. I have some regular servers, possibly Yahoo Gmail, but it could be any other server. Now I’ll try to only receive this network information into my BGP table, which means that I have the information for those ten servers or networks in my BGP table. But I am not going to maintain the complete internet routes for everyone. And if so, to reach those ten routes or ten networks, I’m going to receive the information, and I can run BGP.
And I believe the 10 and 20 dot networks should share the same primary ISP. Alternatively, I can say they will use an ISP route in my LAN. So, in my case, I have four different networks, and I can manipulate the paths so that when these four networks reach these specific servers or networks, they use ISP 1, and the two remaining networks use ISP 2. So I’m going to do some path manipulation based on those specific routes. And if I’m trying to access any other networks other than this, like maybe I’m trying to access some normal Internet traffic or I’m trying to do something other than accessing these specific servers, they just go with the default route. So in this scenario, I’m getting a default route, which means it is reducing the overhead on my router. I’m not going to instal all the routes at the same time. I have some specific routes that allow me to do path manipulation only for specific destinations or networks. at the same time. For the remaining networks, I am going to use the default route that I have.
That’s the second option. Now the third option is exchanging all the internet routes. In the third option, we will not use any default routes; instead, we will exchange everything from the service folder, which means that on my as, any router in my S is connecting to both ISPs, this is my ISP one, and connect to ISP two. I’m receiving all the routes from the service folder, including all the internet routes from this service folder. Now I’ll do path manipulation for each and every network. Any network I go to, I want to access on the internet or any other, as I’ll do my own path manipulation. So based on that path manipulation, it’s going to have the major drawback of being very hard on your routers. Like your routers, they really need a lot of processing. is required to maintain all the route information to run the BGP. And you need to have some high-end devices. So especially this is more appropriate for very big size companies.
Or if you’re continuously communicating, especially with ISPs, I can say that the service board does need to maintain each and every route information that comes in. So if you’re running a specific organization—a big server organization—we can go with the second option. In the case of small companies, we generally don’t use PGP; we simply go with the default route with the primary and secondary options. So these are some of the three different types of implementations for which we use the default routes from the service portal, which is very easy on the routers. Internet traffic is routed through the nearest PGP router. So it’s just forwarded based on your IGP information. Or we can just allow for the selection of some paths, with others falling back to the default route. And the third one is that we are going to exchange all the routes, which is very hard on the resources of the router, but again, it’s going to carry the most direct path or the shortest path. It’s going to carry a shorthand.
Now, depending upon the requirement, you can go with any one of these options. So we discuss two things here. We attempt to comprehend the various methods of connecting to ISPs, including single-home connections, dual-home connections, and the more appropriate use of a default route without BGP. And if you’re running a multi-homed or dual-homed environment, then it’s really recommended to go with BGP. At the same time, we translate some paths. When you connect to BGP, when you’re running, when you’re exchanging routes through the Internet, the ISP, and the service folder, you can have a single default route coming from the service folder, and we’re not going to run BGP here; we’re just getting the default route from the service folder. Or we can have some specific routes from the service folder and some specific routes, and then the remaining will automatically switch back to the default route, and we can have all the routes exchanged from the service folder. So, three different types of implementations So probably we’ll try to jump into our labs, do some basic configurations in our next video. So these are some of the basic things we need to understand before we start getting into BGP configurations.
