On large networks, the size and complexities of routing tables and dynamic routing protocols can require a great deal of work to properly aggregate and distribute routes. In this video, you’ll learn how route aggregation and route redistribution are used to build functional networks in any size environment.
<< Previous: Routing TablesNext: Routing Metrics >>
When you start working with routers, especially on larger networks, you’ll find that your routing table starts getting larger and larger and larger. And one of the things you can do to optimize this routing table is to summarize the routes through something like route aggregation.
You may see this referred to as supernetting, prefix aggregation, route summarization. It’s effectively all the same thing. We’re taking a list of routes and we’re summarizing them down to perhaps one single route that we can then use to make our routing tables much more efficient.
This is going to simplify the methods that the routing table uses to make a decision. So our routing is going to be faster. We’re also going to use less memory since every entry in the routing table takes up space inside of the memory of our router. And if we’re doing dynamic routing there will be fewer routing advertisements that we’re sending out.
To be able to take advantage of this route summarization we’re going to want to either statically route or use a routing protocol that understands CIDR block notation. An example of these might be RIP version two, OSPF, EIGRP, or BGP.
To summarize our routes we would look at all of the routes in our routing table. And we want to know what part of these routes are similar across all of these. The best way to see this is if we take each route and we write it out in binary. So I’ve done that with my routes of 192.168.98.0 is 192.168.98.0. And we’re going to write out all of our addresses, all the way down, for all of the routes inside of our device.
Then we’re going to look at all of these that we’ve written down. And we’re going to try to find the things that are similar and the things that are different between them. And I’ve highlighted this in my list by making them green and orange.
And you can see that you’ve got a similar octet 1. Octet 2 is exactly the same. Octet 3 is a little bit similar. Notice that the first four bits of octet 3 are identical. But the last four bits begin to deviate. So everything from the last four bits of octet 3, and of course everything in octet 4, are going to be different.
So we’ve separated those by colors. And now we can decide where we need to summarise. So we’re going to take this separation that we’ve done. And at the dividing point everything else is going to be a 0. So if we take that binary and we convert it back to decimal we’re able to summarize all of this as 192.168.96.0.
And we need to determine how many bits are going to be identical for all of these. And if we count eight, 16, 17, 18, 19 and 20 bits in total are going to be identical between all of these routes, which means that we could use a mask of /20 to summarise this. So instead of putting all of these separate addresses in a routing table we can simply put one single route in a routing table and point that to the next hop.
Sometimes you have to calculate this aggregation yourself and configure your router manually. But a number of routers and operating systems are able to aggregate these routes automatically inside the operating system. So you may request a list of all the routes that a device knows about. But it is giving you a summarized view so that it’s able to minimize the number of entries in its routing table.
One of the great things about having all of these choices of dynamic routing protocols is that we can find the one that fits perfectly for our organization and use it inside of our company. But other companies may choose to use other dynamic routing protocols. And if we ever need to communicate to one of those other companies what we’ll find is that these different routing protocols will not talk to each other.
For instance, we have RIP in this company. This company’s using EIGRP. This company down here is using RIP Version 2, Company B using OSPF. And connecting them all together is BGP. Our OSPF can’t update the routes that are inside of Company C’s EIGRP. Because these routers are effectively communicating with completely different routing protocols.
So now we’ve got this problem. We have these autonomous systems that are managed by completely different people that have decided on their own sets of dynamic routing protocols. And of course, none of these routing protocols can communicate with each other. You might have RIP Version 2 on one side and you might have OSPF on the other side.
So one of the ways that you can take routes from one system and convert them so that they can be used in a different system is to use something like route redistribution. This can redistribute these routes that we’re learning about via one method and redistribute them through another method entirely.
You might take routes that are being determined through RIP Version 2 and change them so that they can be advertised through OSPF, or take our OSPF routes and redistribute so those can be seen on a BGP network.
Route redistribution also allows us to take routes that aren’t normally advertised on our internal network and advertise them to the outside. This may be useful if we need to advertise a service so that people can gain access to a web server or a file server inside of our organization.