Classless Inter-Domain Routing (CIDR) to Create Subnets Part 1
9 hours 49 minutes
IP addressing or classical IP addressing everything that we've discussed work so far.
What we're going to get now is some of the limits of using a classical addressing system and moving to a classless system.
This is referred to as C i d. Are classless inter domain routing.
This is going to give us the ability to submit our network.
Somebody means I'll take one big network, id the 10 Network and break it into smaller networks. So I have 10.8 10.16 10.21 network. However I choose to submit.
What will allow me to do is create unique network IEDs within the overarching network idea of my organization as a whole.
One reason why I might want to do that is to isolate broadcast traffic.
Routers will isolate broadcast traffic to a specific port.
So if I have the sales domain and they have an application, that journey, it's a lot of broadcasts. I don't want those broadcasts going throughout my network.
I can put them on their own sub net, and their broadcast would be limited there.
I may also want to create unique security domains,
for example, For one, my HR computers because they have a list of payroll information and other sensitive data.
I may have avoid network to prioritize traffic, or I may just want to break them up my network into smaller, easier subnets.
Don't forget what type you have said that masks that separate network portion and host portion.
Hopefully, we've laid the groundwork for what we're going to do now.
You've already said binary ones in the sub net mask indicate a network address,
whereas zero in the sub net mask indicates a host.
So we're certain talk about things in binary hold that thought for just a minute.
The thing that's so rigid about classless addressing is an entire octet is either network or host.
When we see this looking over on the right side here, classical addressing is inherently wasteful.
Why is that?
If you look at the four octaves and see a full octet, is either network support a host support?
Then, if we start with Class A sub net mask, which is the 1st 255.0 point 0.0, I've got three octaves for hosts
that will support 16.7 million hosts. That's a lot
the way we figure that out is each octet has eight series for the whole going back.
The way we figure that out is each octet has eight zeros for the host portion.
The formula is two to the power of X, where X is the number of zeros
you find your value. But then you also have to subtract two because we have to count from the network ID and the broadcast address.
So when we look at two to the 24th power minus two, we come up with about 16.7 million hosts.
We go down a Class B
and we have support for two to the 16th power,
which is 65,000, 356.
We subtract to account for network idea broadcast.
So you have support for 65,354 hosts.
Class C only has one octet
or eight bits to support hosts,
so you have to. The power of six minus two gives us support for 254 hosts.
So when I say this is inherently wasteful,
let's imagine I have 255 hosts.
I can't use a Class C.
I have to go all the way up to Class B, which means I'll waste about 65,000 I p addresses.
That's really wasteful.
When you're talking about trying to manage IP addresses in an enterprise environment,
you don't want a huge range of available addresses.
People could take advantage of those IP addresses that would be on the same network,
and that may be a security concern.
But it's just inefficient.
A lot of this goes back at the time when we purchased a Class C address or a Class B,
if we were a huge organization, we purchased a Class A.
We never want to pay for more than what we need. Same idea today
when I have an Internet presence, I have to pay for my IP addresses that are on the Internet,
which is why it's important to choose the correct number.
We've got a very rigid scheme that's inherently wasteful.
So what we want to assess is how we can make it less wasteful but also divide our networks up into additional substance.
We'll take a basic network ID
10 point x point X point x with the 255.0 point 0.0 sub net mask.
I want to show you the mask in binary,
we have eight binary ones. The next three octaves are eight binary zeros.
In this case, the first octet is already taken.
I think of it this way because the first octet will be reserved for the network ID because it starts with being a Class eight address.
The first octet we don't work with.
And the second Oct. It, however, take note of all the zeros
here and in the rest of the sub net mask.
There is support for 16.7 million hosts, which I have no need for.
I'd like to trade some of those hosts for networks or subnets,
and thinking about this will trade zeros for one in our sub net mask.
And that's how we will create additional subnets.
We're swapping the zeros to ones that are sub net. Masks will change.
We'll be sending bits, so to speak, going back.
What were you splitting bits, so to speak?
Let's take a look at submitting when working with the classical address.
For example, let's say I have a class A mask
which normally is eight binary ones and 24 zeros.
If I want to create additional networks for myself,
trade some host support for network support,
I go to the next bit of the second octet and trade it from 0 to 1.
In binary that becomes 255.128 point 0.0.
What's the purpose of doing this?
I created new networks
every bit I steal. I create two to the power of X networks,
where X is the number of bits stolen.
By changing that first bit from 0 to 1, I created two to the power of one network.
So I created two networks.
Just by modifying the sub net mask and stealing or splitting bits, I can create additional subnets for myself.
What if I wanted to create four additional networks?
I still two more beds or a total of two beds.
If I steal the first two bits, that would be 255.1 night. 2.0 point zero
into the power of X networks will be created
in this case to the power of To will give me four new subjects.
This continues. The more networks I want, the more bits I borrow.
Don't forget that changes your sub net mask, which is why knowing a bit of binary does help