Routing tables can be configured automatically through the use of dynamic routing protocols. In this video, you’ll learn the mechanics of dynamic routing protocols and the most popular dynamic routing protocols.
Static routing requires the network administrator manually configure every route on every router in the organization. If you have three or four routers, that might not be a big problem. But if you have tens or even hundreds of routers, you might want to have a more automated way for configuring your routing tables. One way to do that is through the process of dynamic routing.
Dynamic routing will have the routers handle this process for you, automatically discover these routes, and update each other as to where the best route might be. This means that you as the network administrator don’t have to do any type of static route configuration. You don’t have to SSH into the router, make any configuration changes, or update those routes if anything changes with your network infrastructure. Any time you bring a new router online, all of the other routers will automatically know where the new route is. And if you happen to remove a router from the network, all of the routers also know that that route is now no longer available.
There is a bit of overhead required inside of the router to be able to automatically discover these routes and update the routing tables. This will require some CPU and memory inside of the router, so you may need to do additional monitoring to make sure that your router is able to handle that load. This also requires you to initially configure the dynamic routing protocol. This may be a relatively easy configuration process, or it may require additional planning and engineering to be able to implement it properly in the config of the router.
Let’s look at this network configuration where we have Sam that may want to communicate across the network to Jack or to Teal’c. You’ll notice that there are three different routers. And for this particular example, let’s focus on router one. You’ll notice that router one is directly connected to three different subnets, but there are two subnets on the other side of router two and router three that router one simply can’t see. So we need some way to update router one with those additional routes.
One way to do this would be to have router to send a routing update via EIGRP to router one. Router one receives that update and then updates its own routing table with this new route to 10.10.20.0/24. And the way that you get to that network is to go to 10.10.40.2.
We also have another EIGRP update that has come from router three. And when that update is received, router one identifies a new route to 10.10.30.0. That’s the subnet down at the bottom.
And you can see that the next hop is 10.10.50.2. All of these updates occurred in real time behind the scenes without any type of user intervention. And if we happen to add or remove routers from this network, this routing table will be updated with an additional EIGRP update that will either add or remove those routes from the table.
Although it looks relatively straightforward to perform these dynamic routing updates, there’s actually a lot of work that occurs behind the scenes. First, the router needs to listen to the traffic that is on the local subnet and see if there are other routers that are sending routing updates that it can use to build its own routing table. These are very often sent directly from router to router, either directly or with a multicast.
Once a router then builds its routing table, it needs to inform other routers of routes that it knows. So it will send its own multicast to other routers that are nearby, informing them of all of the routes that that router happens to know. Once the other routers receive this update, they need to interpret this information and determine if that’s a better route than what they already have or if they should use this as a secondary route.
Different routing tables use different methods to make these decisions, and they will update their routing tables differently depending on the protocol you happen to be using. And of course, if there is any change to the network infrastructure, these routers need to inform all of the other routers that the change has occurred. So if we add a new link into our router, we remove a router, or we add a router, we need to make sure that all of our routing tables are updated across the entire network.
There are a number of different dynamic routing protocols to choose from. So which one would be the best for your particular implementation? Different routing protocols make routing decisions in different ways.
For example, would you like your traffic to be routed based on the state of the link– whether the link is up or down– or would you like to be able to make that decision based on the number of hops that a link might be away from you or the speed of that connection? Based on the routing protocol that you’re using, that decision might be very different. Some routing protocols will use a different criteria to determine the best way to get to a remote location. You might also find that some routing protocols are able to make changes very quickly if anything happens to the underlying infrastructure. So if you add a new router to the network, it might take a number of seconds or it might take a minute, depending on the type of dynamic routing protocol that you’re using.
And in some cases, we need to take into account the type of router that we’re using. Some routing protocols such as BGP and OSPF are very common and can be used across many manufacturers’ devices. Some routing protocols such as EIGRP may work best in a Cisco-centric environment. So if you have a lot of Cisco routers, that might be a better routing protocol for you.
Although EIGRP tends to be very Cisco-centric, you may find EIGRP is also available on other manufacturers’ routers. But there are some aspects of EIGRP that are proprietary to Cisco, so you tend to see it mostly in Cisco-related configurations. But one advantage of EIGRP is that it is relatively easy to set up. You would turn on EIGRP in your Cisco router, give it a few minor configuration options, and it’s now up and running with the EIGRP protocol.
When there are changes, EIGRP tends to converge relatively quickly, and it’s also able to identify any loops and prevent those from occurring on your routed network. EIGRP is also very good at identifying other EIGRP enabled routers and sending updates over a minimum of network traffic. This keeps your network more efficient and leaves your bandwidth available for other applications.
If you want a more generic dynamic routing protocol that can be used across different manufacturers’ routers, then you might want to try OSPF. This is the Open Shortest Path First routing protocol, and it’s one that you’ll find available on a number of different manufacturers’ devices. This is often implemented in a network where you have complete control of those systems. We often refer to this as an AS, or an Autonomous System. For example, if you’re running a wide area network that has 50 different routers, those 50 routers would be part of your single autonomous system.
Another advantage of OSPF is that it’s available on many different manufacturers’ devices. This is a common standard, and you can download this standard from the internet and read through every aspect of OSPF. We refer to OSPF as a link-state protocol, which means it determines what the best route is based on uptime and availability between the different OSPF routers.
OSPFs can be used to assign costs to an individual link, so certain links may have a higher cost than another. This allows OSPF to make routing decisions based on what the least cost might be. This cost is often associated with how much throughput may be available on a particular link, if that link is up or down, and how long it takes to traverse that particular connection. With OSPF, the lowest cost and the fastest path is going to be the best route to a remote location. And if there are identical costs on OSPF, many implementations will allow you to load balance across both of those links.
If you need to route traffic outside of your autonomous system to other organizations, then you might want to use an external gateway protocol such as BGP. BGP is the Border Gateway Protocol, and it’s commonly used on our wide area networks and internet connections. Connecting the internet together and being able to dynamically update the routes on the entire internet is a daunting task, and this particular routing protocol was designed with this specific task in mind. Sometimes you’ll hear BGP referred to as the three-napkins protocol because it was sketched out on napkins initially to solve this particular type of problem. If your organization has one or more connections to the internet and you want to be able to dynamically route to those internet sites, then you’ll want to use a protocol such as BGP.