I recently wrote a two-part series about the basics of Cisco IP subnetting ("Cisco IP subnetting 101: Five things you should know" and "Cisco IP subnetting 101: Five more things you should know.") In response, several TechRepublic members posted comments in the article's discussion or contacted me with questions and requests for more advanced information about IP subnetting.
So I decided to take advantage of such requests and use them as fodder for my column. A couple weeks ago, I answered one member's question about all 1s and all 0s subnet masks. This time, let's look at another member's request for more technical information.
TechRepublic member Kevaburg offered the following feedback and requests for more details.
"I think I would have liked to have seen content showing how taking bits from the host to the network portion of the address changes the amount of subnets you have.
"A basic discussion of route summarization and a bit more about CIDR and the roles they play within subnetting would have been more useful than telling us all what we already know."
Changing bits on the subnet mask
Let's start with the first part of this request: How does moving bits from the host to the network portion of the address change the number of subnets? By taking away bits from the network portion of the address, we reduce the number of subnets and increase the number of hosts.
This is always the case with a subnet mask. Adding 1s means increasing subnets and decreasing hosts per subnet. Removing 1s means decreasing subnets and increasing hosts per subnet.
Let's look at an example. Say we're starting with an IP network of 126.96.36.199 and a subnet mask of 255.255.255.0. In binary form, the current subnet mask looks like this:
11111111 11111111 11111111 00000000
The sequences of 1s represent the network portion of this IP address, and the 0s stand for the node or host portion. To keep the example simple, let's take eight bits from the network portion of the address. So, we remove the last eight 1s and change them to 0s. The new subnet mask would look like this:
11111111 11111111 00000000 00000000
With the original subnet mask, we had 254 useable hosts in the network (which we can determine by using the hosts formula: 28-2 hosts). With the new subnet mask—which is now 255.255.0.0—we have 65,534 useable hosts in the network (216-2).
We increased the number of hosts dramatically, but we also decreased the number of subnets by 256, which we can determine by using the networks formula: 28. (If we're using the original networks formula, 28-2, we would decrease the number of subnets by 254.)
So what are you going to do with a network that has more than 65,000 hosts? Actually, you're not going to use all 65,000 hosts.
You could assign it and address it, but at some point, your network will start to bog down with all the network broadcast traffic—and it will eventually grind to a halt. In reality, you would more likely use this reduced subnet mask to represent a supernetted network, which segues nicely into the second part of this member's request.
Supernetting a network
Supernetting (also known as route summarization or route aggregation) uses classless interdomain routing (CIDR) to address a number of IP subnets with a single route. We call that single route a summarized route or a supernet (in other words, the inverse of a subnet).
To learn how to supernet a network, let's look at another example. Let's say we have four IP subnets on the four LAN interfaces of our router: 188.8.131.52/24, 184.108.40.206/24, 220.127.116.11/24, and 18.104.22.168/24. We want to summarize these networks into a single route that we can advertise across the WAN, which reduces the number of routes in the remote routers.
We could summarize these routes with this supernet IP address: 22.214.171.124/22. This single IP address references all four of the IP subnets. Here's a look at it in binary form:
IP address: 00000001 00000001 00000000 0000000
Supernet subnet mask: 11111111 11111111 11111100 0000000
Notice the third octet of the supernet subnet mask: 11111100. This allows the last two bits of the third octet to be any combination of 00, 01, 10, or 11. So when advertised, this supernet mask would show that any of the four subnets are available from the router.
Keep in mind that when subnetting or supernetting from the classful boundaries, you must use a routing protocol that supports variable length subnet masks (VLSM) and CIDR. Your options include Routing Information Protocol version 2 (RIPv2), Enhanced Interior Gateway Routing Protocol (EIGRP), the Open Shortest Path First (OSPF) protocol, and Border Gateway Protocol (BGP).
Of these protocols, EIGRP is the only one that summarizes at classful network boundaries by default—a capability that you can turn on or off. On the other hand, OSPF requires manually entering a summary route with the summary-address command. BGP disables autosummary by default, but you can turn it on, or you can use the aggregate-address command to create your own summary route.
Whether you call it route summarization, route aggregation, or supernetting, this practice is essential on the Internet. If every carrier advertised every specific route it has, it would overwhelm the memory of the Internet BGP routers. For example, my company's BGP Internet router has 125,000 routes to Internet networks, and most of these routes are supernets. However, because the advertising routers summarize their routes, the router is able to receive all Internet routes using only 125,000 entries.
Miss a column?
Check out the Cisco Routers and Switches Archive, and catch up on David Davis' most recent columns.
Want to learn more about router and switch management? Automatically sign up for our free Cisco Routers and Switches newsletter, delivered each Friday!
David Davis has worked in the IT industry for 12 years and holds several certifications, including CCIE, MCSE+I, CISSP, CCNA, CCDA, and CCNP. He currently manages a group of systems/network administrators for a privately owned retail company and performs networking/systems consulting on a part-time basis.