Quantum Networking (Part 2)

Welcome back. We've discussed schemes used for managing IP addresses

which naturally leads us to the concept of IP subnets.

We want to break off certain machines and put them into groups that we can talk to you at once.

In order to facilitate that we add what's called a subnet mask

which is a way to mask off a number of machines that will receive a certain broadcast.

When expressing subnets, we have the same concept as an IP address, 4 octets.

To give the simplest explanation,

a 255 in an octet means that the corresponding octet of the device IP will have to match to make that group.

A zero in an octet means a device can have any IP address in that place and be a member of the group,

as long as it meets the other conditions.

Values other than 255 and 0 are allowed in subnets, but we'll discuss those later.

So let's look at a common case.

Here is a subnet that includes the IP address 192.168.1.101

with a subnet mask of 255.255.255.0.

We know that all machines on this subnet must have 192.168.1

in the address due to the 255s in those corresponding places in the subnet mask.

We have a 0 in the last place, which would suggest that 0 through 255

would all be allowable values for an IP address to have in its last place and still be included in this subnet.

However, there are two special cases. The first address in the subnet range (0 in this example) is reserved to define the subnet.

And the last address is reserved as the broadcast address for that subnet.

So that leaves us with the following:

A subnet address of 192.168.1.0 , which defines the subnet.

A host range of 192.168.1.1-254, which are IP addresses that machines on the subnet can have.

And a broadcast address of 192.168.1.255;

that’s the address we can use to talk to all of these machines at once.

When subnets were originally specified, they were what we refer to now as ‘classful subnets’.

Each range was specifically defined.

A lot of the IP schemes we see now are holdovers from those original definitions.

Class A networks are defined as having first octets 1-126 with a subnet mask of 255.0.0.0.

The most commonly used subnet in the range is 10.0.0.0.

Note these subnets have up to almost 16.8 million possible hosts.

That’s never done these days, so if you’ve seen it, it’s mostly in the form of a more restrictive subnet.

Class B networks use the subnet mask 255.255.0.0 with the first octets of 128-191.

This allows for a little over 65,000 hosts per subnet, which is still a large number of devices.

The most common address schemes coming from this class are ones beginning in ‘172’.

Class C networks are the most common ones we see,

from your own home networks to those in our workplaces, etc.

These are small subnets with up to 254 devices.

Almost everyone has seen a 192.168.0.x or 192.168.1.x IP address if they’re looking.

And now that routers are so much less expensive and prevalent, subnets tend to be in this range or smaller.

There’s no need for a giant class A or class B subnet anymore.

Note that the IP ranges in the Class A, B and C network definitions are NEVER found on the internet.

These ranges are specifically set aside for internal networks.

Addresses of machines on the internet themselves

are very carefully set aside and assigned to avoid catastrophe.

We mentioned before that subnets could be defined outside of the rules

and classifications we just discussed.

Doing this is referred to as Classless Inter-domain Routing, or CIDR.

To explain, we’ll need to convert an IP address to binary… which is even more fun than hexadecimal.

Let’s say we wanted to start with IP address 192.168.1.1 and create a subnet smaller than class C.

The binary representation of that address is to the right.

Note a class C subnet would be represented as 24 ‘1’s from the right, with 8 ‘1’s representing 255 as a decimal.

To make our subnet one size smaller, we add a 25th ‘1’ to the subnet mask going from the right.

Every 1 in the subnet mask represents the bits in the address that define the network, so they’re called the ‘network’ bits.

The hostrange of the subnet is determined by the 0s in the subnet mask (represented here in dark blue).

If we convert all of this back to decimal we have a subnet mask of 255.255.255.128.

In what we call CIDR notation we specify the subnet address,

which is 192.168.1.0 with a slash and then the number of network bits, which is 25 in this case.

This makes the allowable hostrange 192.168.1.1 to 192.168.1.126.

The final address, 192.168.1.127 that becomes the broadcast address for the subnet.

So, does your head hurt yet? 'Cuz mine does!

You know, all of this seems complicated, but there are online calculators to make this easy.

Here are a couple of examples.

If you’re given an IP and netmask by corporate IT, you can easily find the host range.

If you need to define a custom subnet, you know can do that as well.

So, let’s take a break here, and in the next section we’ll dive into how to actually troubleshoot

some of these issues when you’re on a project site.