New to Busy?

Course: IP Version 4 Addressing And Subnetting Deep Dive (Part 7)


7 months ago7 min read

Hello and welcome, I am currently working on a video course 'addressing' IP version 4, and I've decided to post my slides and scripts here on STEEM as I go along. I appreciate any feedback or suggestions, or if you are new to the world of computer networking, I hope you enjoy them!

Here is part 7 of a multi-part series. How many parts will there be? I have no idea because I am posting them as I go along. In this part we will continue to discuss subnetting and end with a subnetting case study.

If you missed any sections, click one of the links below for your favorite front end:


From here we are going to work with a /16 as our reference point, this is a full Class B worth of addresses. Rather than dealing with fractions of a /24 as we did in the previous section, we will be working with multiples of /24s. To do this, we simply add a bit to the subnet mask.

Note: A few times in my career I’ve seen a a single LAN network with a /23 or /22, but it’s not really something I recommend because the more IPs you have on a LAN, the less efficient it becomes. If you did decide to use a large block such as this, you would still need to account for the network and broadcast address, so 2^n-2 would still apply there. You can see in the table the number of possible hosts is excessive for anything larger than a /24.

More than likely if you are dealing with a network larger than a /24, your intent will be to divide it up again as part of a hierarchical addressing scheme. More on hierarchical addressing later.

As you scan through the table, you can see what happens to the number of possible /24 subnets and number of possible hosts. As with the /24 subnetting, we are adding a bit to the subnet mask dedicated to the subnet address, which cuts the number of hosts and networks in half. The change in subnet mask can be seen in both decimal and binary.


As we continue down the line, we keep adding a bit, and halving the possible number of hosts and /24 networks, until ultimately we’ve added 8 bits and have created a single /24.


Unless you are dealing with a a RFC 1918 private network, it is unlikely you will ever need to perform subnetting on anything larger than a /16, unless you work for a regional Internet registry, Internet service provider, or government entity. Since there are plenty of enterprises that use hierarchical addressing to break up a private network, you should be familiar with how to do it. There are also applications for creating summary addresses and default routes, which we will get into later.

Using a /8 as a reference, we can measure its size in terms of how many /16 networks it can be divided up in, which will also divide the possible /16 networks and hosts by half as we add bits.

Note: In this case, I’m not going to bother subtracting a possible network and broadcast, because you should never see a single network of this size.


Continuing on, we get the same result. We add a bit to the subnet mask and the number of /16s and possible hosts addresses are cut in half, until we reach a standard 16-bit class B address.

You could also subnet with less bits than a /8, but it is extremely unlikely you will ever encounter this. If you do, the principles are the same, you are merely dealing with multiples of /8 networks.


Now that we’ve got subnetting down dividing a larger network down into equal parts, we can look at a more realistic scenario. This involves taking a larger network and subnetting it down to heterogeneous, or different sized subnets. For our example, let’s say a small enterprise division has been allocated The division requires different IP addresses for different purposes. For example, you could say the /25 could be used for computers, the /26 could be used for IP phones, the /27 could be used for printers, and so on. This is all accomplished by manipulating the number of bits in the subnet mask.


Another example is we are assigned a /22, which can be divided into a single /23, and two /24s to start. The /23 could represent a headquarters office, and each of the /24s could represent a branch office. From there these allocations could be subnetted down an additional time to form additional hierarchy.


Widgets INC has two locations, Chicago and New York City. You have been tasked with assigning IP addresses to the PCs, IP Phones, and Printers at each location. You do not have to worry about any other addressing requirements such infrastructure or management.

  • Chicago has 25PCs, 18 IP Phones, and 5 Printers
  • New York has 22 PCs, 17 IP Phones, and 3 printers
  • Neither location anticipates more than 10% growth.

What is the minimum contiguous IP address block that you must request from your Internet Service Provider? Let’s take a look.


To begin with, let’s account for possible growth, which will never be more than 10%. Just calculate 10% of the immediate requirement and add it to the total. Always round up and add an additional IP address for the default gateway!

You can see after we account for future growth, the two site’s requirements are very close. This means that we can standardize the subnet allocation for each one. At this point you can determine the smallest possible subnet that will account for the requirements. For both PCs and Phones that is a /27, which can accommodate 30 IP addresses. So we will require a total of 4 /27s. For the printers, we will require /28s, because the a /29 cannot accommodate the 6 printers along with the default gateway. If you’re still not clear on what a default gateway is, that’s okay.


We know that a /27 represents 32 IP addresses, and a /28 represents 16 IP address. If we add up the total requirements for each site, we get 80. This represents a grand total of 160 IP addresses that we need.

Given what we know about subnet sizes, at 128 IP addresses, a /25 is the smallest possible subnet we could use for each site. At 64 addresses, a /26 clearly would not suffice.

We can combine two /25 networks into a /24, so we decide to request a /24 network range from the ISP. The ISP provides us with the network address Let’s break that down into subnets.


Here we have the subnet breakdown. Each site has their own /25, and we have broken those down into appropriately sized subnets. This leaves us with the spare IPs for Chicago, and - 255 for New York. We have over a /27 worth of IP addresses left over. We could either decide to hold those in reserve, or we could adjust our subnets and expand their size. For our purposes, we will hold these extra IPs in reserve. You never know when a new technical requirement might come up.

Thanks for viewing. In the next part, we will learn how to determine if two subnetted addresses exist on the same network, we will also learn about supernets, summary addresses, and wildcard masks.


Sort byBest