Cisco BGP Essential Training

- As we begin our study of BGP, I'm starting with a quote from one of my favorite books, one of my favorite movies and one of my favorite tv shows ever. I will spare you my Charles Kingsfield impersonation, but when he says "the study of law is something new and unfamiliar to most of you unlike any other schooling that you've ever known before". The same can really be said for BGP. And here's the entire deal with BGP. It is unlike anything you've studied to this point. So if you haven't seen this before, don't expect to absorb it all the first time you see it in action.

The rules are different, the concepts are different, the terminology's different. Just about everything is different from the IGPs that you've studied to this point like OSPF and EIGRP. Now it throws a lot of people for a loop the first time they see it and that included me. And the key to initial BGP success and really all BGP success is to take one concept at a time and then one attribute at a time and before you know it, you've mastered the fundamentals of BGP. And what I'm going to do in this part of the course, which is the largest part of the course is when we get to the attributes which will be soon, I'm not going to just hit you with 12 of them and then we're going to go back in the video and look at 'em.

You're going to see every single one of these attributes in action. And you'll see exactly what they do. So it's some good stuff here. It's just new stuff for many of you. And for those of you with one eye on the CCIE down the road, this is some of the most important study you'll do. Because, of course it's true of anything, if you don't have the fundamentals down, you can't really master the advanced skills. But with BGP, as you'll see, it is particularly true. Now the first natural question is, well, if BGP is so different, if it's not really like EIGRP or OSPF, you know, what the heck is it? Well here's one definition for you: An internet protocol that allows groups of routers called autonomous systems to share routing information so that efficient loop free routes can be established.

BGP is commonly used within and between our internet service providers; the ISPs. Now that definition started sounding a little bit like EIGRP, right? I mean, as soon as you hear autonomous systems, you thing EIGRP. Well, both EIGRP and BGP use autonomous systems. We're actually going to be typing in some AS numbers here in our BGP labs as well as we did in our EIGRP labs. Both protocols use that term to refer to a logical group of routers. And the difference comes in really where BGP fits in the big scheme of things.

Because BGP is an exterior gateway protocol that runs between ISPs where EIGRP is an interior gateway protocol that runs in the networks connected by the ISPs. I've always put it this way, BGP runs on top of OSPF and EIGRP. It runs higher in the big scheme of things. Now BGP also shares some other characteristics with EIGRP. BGP supports VLSM and summarization. And we'll do some summarization here in one of the labs.

BGP sends full updates when routers initially become neighbors, or in BGP speak, really peers. And after that, BGP will only send partial updates reflecting the latest network changes. Because, BGP updates can literally contain tens of thousands of entries. And in that case, we really don't want those going out, say, like RIP does every 30 seconds and then send the full table. That would not be efficient. Now BGP, as I've mentioned, we have peers. We have neighbor relationships; the same thing.

They create these relationships before they exchange routes. And there keepalives that are sent to keep the adjacencies alive. And, of course, we know the deal here, if those keepalives disappear, the adjacency is gone. Now, BGP neighbors exchange much more information, much more extensive information, I should say, about networks than our IGPs do. Now the additional BGP path information comes in form of those attributes that I mentioned a few moments ago. And these path attributes are contained in the updates sent by BGP routers.

And we're going to master these attributes; especially the fundamental ones because they are the key to success with BGP. And it also starts with knowing which ones are, what we call well known and which ones are optional. And if these are terms you've heard before, we're going to go over those lists. Believe me, if you haven't heard 'em, don't worry about it because we're going to get to it in the next video or two here. We got a little falderal to go through before we get to the labs so a few rules we need to know. And right now we're going to take a quick look at some general Cisco best practices regarding when BGP should and should not be used.

Now, when BGP should be used is when your company is connecting to more than one AS or ISP because decisions on the best links to use can be made by utilizing these BGP path attributes. We will see that live. The routing policy of your organization and your ISP just may differ a little bit; that's when BGP can come in handy. And if your company is an ISP to begin with because traffic from other autonomous systems use your AS as a transit domain, BGP will definitely come in handy.

So in short, if your AS has more than one connection to other Ass or other ASs are using your AS as a transit domain, you'll definitely be using BGP. There are a few circumstances when you should not use BGP. When there's a single connection to the internet or to another AS and no redundant link exists. When you don't care which path is used to reach a route in another AS you may have that situation. And when router resources are limited; again, as always, that's a relative term that BGP does hammer the memory and the CPU quite a bit.

And also, when there's a low-bandwidth connection between multiple ASs. In that situation, we may go with some static and default routing instead of running BGP. Now all of this will come a little more clear, first off when you watch the video more than once. But secondly, when you see BGP in action. And at the beginning of the next video. Coming up next, we're going to get our first BGP adjacency built.