Answer (Part 1): Subnet Design Exercise

Today’s post shows the beginning of a solution to the subnet design exercise I posted last week. This post isn’t all that meaningful without reading the other one first – after that, and after you take a crack at it of course, dive in!

The figure shows five subnets with varying requirements for a number of hosts. However, the second requirement stated that you should use a single mask throughout the network, so you can ignore the number of hosts per subnet for the smaller requirements, and just pay attention to the largest requirement: 300 hosts/subnet for R1’s LAN subnet.

To support 300 hosts, you need H host bits such that 2^H – 2 => 300. Some quick math tells us that H=8 is too few (2^8 – 2 = 254, but 2^9 – 2 = 510). The second requirement also told us to use the fewest host bits, so while H=10, H=11, and so on would work, H=9 is the lowest value that meets the requirements. The mask that has 9 host bits is /23, because /23 by definition means 23 prefix bits, leaving 9 more bits to round out the 32 bits in an IPv4 address.

The next step is to list all the subnet IDs, or at least enough to see the pattern. I leave the how/why as an exercise, but here’s a partial list:

172.16.0.0
172.16.2.0
172.16.4.0
172.16.6.0
.
.
.
172.16.250.0
172.16.252.0
172.16.254.0

I’ll stop there for now, and list the rest of the answer in another post, in case some of you who’re following along in real time just needed a little nudge to get to this point. I’ll post the rest in a few days.