All language subtitles for 05 - Cisco Foundations - Basic IP Addressing - Filling in the Gaps-eng

>> If I were to revise the title of the last
nugget, it would be "Basic IP Addressing,

the Essence of Networking"
would have been a great title.

But it's too late, it's gone because really,
in the last nugget, we saw the essence

of communication from one host to another,
how a host knows when to send things

to its default gateway and get
off of its network and all that.

So, you get the real feel of addressing
and how networking communicates

or how networks communicate using TCP/IP.

However, there's just a lot of details and it
was almost, if you remember in the last nugget,

it was almost awkward when I
said, "Oh, by the way, let's pause

and talk about the classes of addresses."

Just because it was such a, such
a good role that we were on,

it's like here's how it all works, that
it didn't make sense to stop and, "Oh,

by the way, there's this detail."

So, that's why I created this nugget
and I felt what a great title.

Basic IP Addressing, Filling In the Gaps.

You saw how it all works.

It's like-- it's like jumping in the Ferrari
and hitting it going, "Yeah, that was awesome,"

you know, and you reach the finish line and then
somebody runs up and is like, "Oh, by the way,

you needed gas to do that,
you know, in your car."

And by the way, we-- you
know, we changed the oil.

You're like-- like there's all these
details you never even thought about.

You were just kind of like, "Man, my
hair is blown back, that was amazing."

So that's what I want to do now.

I want to fill in the gaps and talk
about kind of the technical details

that makes everything in the last nugget happen.

The first detail I want to fill in is
the addressing, as in how did the device,

the network device get the address.

Well, there's many different ways
that it could have gotten there.

One of the ways, and I would say, this is
probably used for servers and key devices

like printers and routers, you know, it's where
you get involved and statically assign it.

This is an old school picture, I think
it's like Windows 95 or something

that the network properties look like that.

We got Windows 7 here.

Let me go to my network status.

Actually, I should show you this.

You know, go to the control panel.

You know Windows 7 kind of hides everything.

What I usually do is just type
in network and jump straight

to view network connections, it's what I do.

But I mean you can poke your way
around and get there another way.

So, I've got 2 adaptors on here, one

of them is my LAN connection,
it's my physical network card.

The other is VirtualBox.

VirtualBox, it's actually on
this spiffy CBT Nuggets machine.

It's a way of doing virtual
machines on your computer for free.

It's amazing.

You know, it's one of those VMware
replacements or virtual PC replacements

because Oracle wanted to get into that mix.

So that, yeah, I mean ignore,
that's not a real network adaptor.

And what I usually do, if I'm on my computer
and this is just me 'cause that's what I do.

I say, "Okay, well this is,
this is my LAN connection,

you know, WAN or something like that."

Or if I know what kind of network card is in
there, I'd say, you know, this is my Intel,

you know, da da da da da
network card LAN and like I put--

I put a nice good description that
makes sense to me in there and I do

that on things like servers and servers.

You know, I don't do it on my workstation
too much just because servers a lot

of times will have more than
one network connection.

So I right click on this guy, go to
properties, got all the protocols.

I open the properties for TCP/IP version 4 and
I see right now my recording machine is set

up to obtain an IP address
automatically, that's DHCP.

But I could come in here and type in I want
this guy to be 122.30.100 and well, you know,

use an address between those ranges and put
my subnet masking and all that, give it a--

I could-- you see what I mean,
I could statically do this.

And the reason I would do this is for machines

that I wouldn't want the
IP address to change on.

See, the alternatives is statically typing it in
because this, this becomes if you imagine this

and scale it to hundreds of computers in your
network, it becomes a nightmare because you have

to go around, and number one, type them all in.

But what happens if a machine gets
replaced and somebody forgot it

and then they type in a duplicate.

And now you get errors across them now.

I mean static addressing does not scale
to all the computers in your network.

So, let's just use for key devices.

What most of them will use is DHCP, the
Dynamic Host Configuration Protocol.

And what that is is where I have my switch
right here, computer plugged into it.

This goes to a lot of times a Microsoft
server like a DHCP server where I would go in

and configure a pool of addresses on here.

Just trying to think if there's
a server I can show you.

I found one.

Too many people can just bust the
server out of their back pocket.

So this is a server that we have running here.

And I went in there, it's
actually already a DHCP server,

but for example if this were a brand new
server, I could come in here and say, you know,

I want to do a new scope, a new scope
of addresses to hand out and I want

to say this is my local network or whatever
description you want to put on there.

Starting I want to hand out
122.30.100.50 all the way

up to 122.30.100 dot let's say 70, you know.

So it's what subnet mask do you want on there.

And I'll use 255 255 255 0, and you can see
you go through and say, well, what exclusions.

Is there ones that you don't
want to put on there?

How long should you want to hand
this out for, what is the duration

of the lease before the client has to
give it back if they're not using it?

If they are using it, they'll get
to give, it's not like, you know,

8 days and bam, you're cut off from the network.

You know, you have the options like do
you want to give them a default gateway.

It's really-- I mean it gives you the
feel to where you go, okay, this is,

this is really a dynamic pool of addresses
that I set up, you know, here's my pool.

And as the computers boot, they actually
send a broadcast message saying, "Hello,

I need an IP address," and all the different
devices on that network that are showing there

as well get that message 'cause it's a
broadcast as well as the DHCP server.

And it goes, "Oh, fantastic.

Let me hand you one from my
pool and you can use this

for however long you've configured
that lease time."

So DHCP, again, I really want
to emphasize it's broadcast base

because that should already be
planting in your mind that, okay,

this doesn't go past the router, right?

There's only is on the LAN and how
that works, so a lot of people do that.

Now, you can set up a DHCP
server unlike a Microsoft server,

that's what I just showed you
with Windows 2008 R2 server.

But you can also use for instance a
Cisco router can be a DHCP server.

You can even buy if you're company
is big enough a dedicated appliance

that all it does is DHCP services
'cause, you know, DHCP is so critical.

If it goes down, you know, your
clients fall off the network, so.

So DHCP allows you to configure things
automatically and then the clients are just

as easy as saying, "I just want to, you
know, hit that bullet, which is the default

of obtaining an IP address automatically."

Now, it's funny I was, you know, I was just
like, oh, how else can they get the IP address.

This concept that I just underlined right there,
DHCP relay is really a CCNP level concept.

They don't really talk about
it much at the CCNA level.

But it's funny, I'm sitting here, you know,
what are we, four, four nuggets into the series

at this point and I'm going,
I think you guys can get this.

I think with this stuff that we talked about, in
the series up till now, you guys would get this.

Meaning, let's say you have an organization
and for now I'm just going to use routers

but I'm going to expand your mind
later to a concept called VLANs.

But, you know, let's say you've
got a whole bunch of small offices,

right, and that's it, that's right.

So I've got no office.

I'll just draw a switch here,
office A down here in Arizona.

I've got office B in Texas,
office C in California.

So we've got-- you know this comes
down to a switch and we've got all

of our clients and all that kind of stuff.

And maybe, maybe these are little insurance
offices spread all around the world

and you don't really want to buy a DHCP server.

You don't want to buy a server to have
to plug in 'cause that's expensive.

I mean you got two people at each office.

We don't-- we don't want to do that.

Now, if this were my environment, I would
say, well, I would just make the router,

the DHCP server then and have
it hand out the IP addresses.

But just for the sake of the
argument, we don't want to do that.

So all of these guys connect you the
main router, well say in Michigan,

and this is where your corporate
network is where you've got, you know,

tons and tons of switches and servers and
hundreds of users that work out of this office

and right here, you wanted to run DHCP
services from a central point, meaning when I--

you know, I have a client up here that ask for
an IP address will come right here to the server

and get it, but I want to set it
up in such a way that these guys

at the remote offices also can get their
IP address from that server up there.

Well, that's a problem.

Why? Who can follow why that's a problem?

It's all right here.

DHCP is based on broadcast.

This guy boots up.

He's like, I don't have an IP address,
hello network, can somebody help me out?

Give me an IP address.

And what do routers do?

[Inaudible] That's the sound of a
broadcast dying if you're curious.

The broadcast has stopped right there, it does
not go any further unless you enable service.

And the reason I want to bring
it up is it's everywhere.

You see, DHCP relay all over the place, so I
want you guys to know this at least exists.

You can go in to your router right here if
it's a Cisco router and enable DHCP relay

which takes a broadcast message
if it's a DHCP request.

And you see, it's not all-- I mean
broadcast happen all the time.

It would be a mess if it was all of them.

It says, I'm going to look for DHCP request and
I'm going to send that, you know, take that--

you know, it's a-- it's a DHCP broadcast and
I'm going to package it as a unicast, you know,

put all the source and destination
headers that we've talked about on there

and send it directly to the DHCP server.

And the DHCP server can say, "Oh okay,
well I've got an IP address for you

and let me hand it right back down."

Well, the reason they call it
DHCP relay, there is the keyword.

That actually hands it to the router
and the router responds to the client

as if it was giving it its IP address,

which is amazing because I can then
centralize all my DHCP services at one spot

or in a really high priced cluster of servers

that maybe are very redundant
and all, all of that.

So, that is one example of DHCP relay, but
once we explore different technologies,

especially VLANs, I'm going to
show you, man, there's so much more

that DHCP relay can be used
for on that side of thing, so.

So that, that is another way that
somebody could get an IP address.

Last one is multiple IP addresses.

There are times, mainly on servers,
where you might want to have something

to have multiple IP addresses or even
multi-homed addresses, meaning, maybe I've got--

and again, without fully discussing VLANs
is this concept gets a little tough,

but maybe I've got two different
portions of my network at the office.

This is our central office
so it's getting messy.

And, you know, maybe this is the public
portion of my office where, you know,

it's kind of like the lobby, I've got
public Wi-Fi and all that kind of stuff.

And then I've got the private
portion of my office.

And there are two different switches connected
to two different interfaces on a router.

They're two different networks.

Well, what I can do is I can
plug in and interfaces, you know,

two interfaces on that server, and give it
an IP address in each one of those networks.

And now that is considered a
multi-homed server, that's multi-homing.

But you can also give a server two
IP addresses within the same network.

Maybe you're running an e-mail
server on here, on one IP address,

and you want a totally different
IP address for handling website.

You know, you run a website.

So, all you have to do, you can even
do this on a computer if you wanted to,

go into the network adaptor, same kind of place,
and you know that I can specify my IP address.

This will be my primary IP address
that I use for everything right there.

But if I want to, I can go into the advanced and
add in as many secondary addresses I want to.

"Oh, by the way this guy is
also, you know, such and such.

Oh, and by the way, this guy
is also, you know, whatever."

You see what I'm doing here, I can go in

and add all these different IP
addresses that this guy can speak from.

Now there's only one primary, as in

when this guy initiates traffic it's
always going to come from this IP address.

However, it will also respond
if it gets a request on any

of these other IP addresses
that I want to give it.

So, I can go in and assign a
device multiple IP addresses.

No problem.

Okay, let's talk about IP addressing now,
specifically, public versus private addresses

and the other special addresses
that exist out there.

Think of the internet.

Most people just think of it as
like, you know, surfing the web

or like a picture of a cloud comes to mind.

But really the internet is a network of
networks, it's the biggest network in the world

that is just one network attached
to another attached to another.

It started with some college
universities wanting to tie together.

They called it the ARPANET.

And before long, you know, people are on
there buying t-shirts and everything else

that you can get online and
books, and whatnot, so.

So, we have this gigantic network of
networks where everything is combined

so everything that's connected to the
internet needs a unique IP address.

So, you know, like here's-- here's,
you know, your computer at home.

You know, it's-- we always draw it as like
a cloud, but really there is a router there

of your ISP and they have
to temporarily give you,

and they do it via DHCP unless you
pay them for a static IP address.

They'll give you an IP address that
allows you to access the internet as long

as you are staying connected, as long as your
device is powered on and actively running.

And you are consuming one of the 4
billion 200 something million addresses

that exists in the world today.

They're-- in the IPv4 address
space, there is a--

actually hang on, let me pull
it up, there you go, I got it.

Its 4,294,967,294 IP addresses,
and you can Google that.

You don't have to know it or write it down.

So, 4 billion some addresses that are there.

Now it sounds like, wow, it's a lot of IP
addresses, but then you start realizing

that we have ultra bazillion network devices
out there that if every network device

in the world got its own public
IP address on the internet,

we would long since have run
out, like it just wouldn't work.

I mean if you think of yourself, you'd have a
cellphone that is connected to the internet,

you have multiple computers at
home, at work a laptop, you know.

I probably have-- I'm just
looking around my office right now.

I see a couple of printers, I see a router,
I see some switches, I have 1, 2, 3,

4 computers surrounding me that
are actively on and being used.

I have my iPhone, I have some IP phone
sitting on my desk that are online.

So I mean, it's just I-- and
I'm-- I'm one man, Jeremy Cioara.

I have all of this stuff and
so it's just not scalable.

So they came up with this idea of private IP
addresses that allows people to assign addresses

that are not usable on the internet.

Now, there's a lot of misnomers
about that that I want to demystify,

but let me first off talk
about the way this works.

What you can do is inside of your
company, or inside of your house,

or whatever environment you're in, let's go
with the house 'cause that's nice and simple

and you just exchange bedrooms for
cubicles and you have an office environment.

You've got your house which, you know,
I just described in my house, I've got,

you know, 4 computers just surrounding me.

You've got a Wii, you've got a-- you've got all,
all of the different things that exist there.

Maybe you have, you know, 20 network
devices in your house which I would say is,

you know, probably about average nowadays.

So, what you can do is just make
up IP addresses in your house.

You know, go ahead and if you're like 90
percent of the homes around the world,

you're probably using 192.168.0 dot
something addresses inside of there

or dot 1 dot something addresses.

And the reason I would know that is
because every NETGEAR, D-Link, you know,

no name device that you go buy
and plug in at your house hands

out these addresses via DHCP by default.

Now, your house uses those addresses, your
neighbor's house uses those addresses,

there is the neighbor house-- neighbor's house
uses those addresses and yet there's no problem.

How does that work?

I mean, I told you every single network
device has to have its own IP address,

so how does my computer have the same IP address
as my neighbor's computer and nobody screaming?

Well, the way it works is
through a system of NAT.

Let's, let's say I've got a router, these
are routers, at the corner of my house,

that is plugged into Cox
or Qwest or whatever ISP,

whatever carrier you're using
to connect to the internet.

This device runs something called
NAT, Network Address Translation,

which allows you to take
all of these IP addresses.

It could be 20, it could be 200,
it could be 2000 IP addresses

and now you would definitely
outdo me maybe if you had 2000.

But 2000-- and it would actually translate
it or some people call it overloading

or some other people call it PAT.

There's many different names for it, and we'll--
again, we'll totally unpack this one later,

translate it to one public IP address that
your internet service provider lets you use.

So maybe they give you the
public address of 5.5.5.5.

Essentially, anything not
shown here or in these ranges

down here are considered public addresses.

So 5.5.5.5, what my router will do is allow
all of these 20 devices to share and be hidden

and pretend to be that one IP address.

And again, I'll show you
how all that works later on.

But my neighbor, they might get 5.5.5.6,
so their 20 devices are never seen

because they're using duplicate addresses
and they all go out as this one address.

So private addresses are not
only useful in houses, well,

they are also useful inside
of organizations as well.

You know, when you're in a large
company, and I'll pick Boeing or Intel

or Nabisco making Fig Newtons, you've
got large manufacturing plants,

you can use private addressing all throughout
your company because that works A okay.

It's just once you want to get to the internet,
the internet will block those addresses

from being used, so you have to NAT
them out, so private addressing.

It's kind of like think of this way.

Remember when-- what was that movie?

1980s there's a movie called-- I just
remember Christian Slater was in it.

Oh, "Pump Up the Volume" where he ran
his own little private radio station

out of his dorm room.

And in the end he ends up getting busted
because the FCC comes in, it's like,

"You can't run a radio station because
those are restricted frequencies.

We manage all the frequencies."

Think of that as like that's
the public IP addresses.

You can't just go out there and start saying,
"I want these, I want those, I want those,"

'cause they probably belong
to somebody else in the world.

However, the FCC has granted broadcast
privileges on some frequencies

which probably sound familiar to you,
like 900 megahertz or 2.4 gigahertz

or 5 gigahertz frequency spectrum.

What are those?

Well, those are used for like cordless phones
or wireless Wi-Fi devices inside of a network.

The reason they call these unmanaged or
unlicensed frequency is because they're like,

hey, you can run whatever you want as
long as you're not maliciously trying

to destroy somebody or somebody else
trying to use those frequencies with it.

And they put, you know, some restrictions
of the power you can use and all that.

But otherwise, anytime we bought a cordless
phone or a new wireless network widget

for our house, we'd have to
go register that with the FCC.

Think of these as like you're
unregistered, your unmanaged IP addresses

that you can use wherever you
want inside of your organization.

You don't have to register or pay for
them at all, they're free for all.

So, that's awesome because it lets us set
up our networks for free without, well--

but, you know, besides all
the Cisco gear we have to buy.

But, I mean, we don't have to think about
buying IP addresses for our network.

Then you drop down here, we
have automatic addressing.

This is a special range of addresses,
169.254, that a computer will sometimes,

depends on what features that computer
has, but sometimes will generate

if it can't contact the DHCP server.

So, let's look back at that picture
or I'll just bring it up here.

This computer is set to get an
IP address automatically, right?

So it doesn't have one at all.

So when it's booting up it's, you know, let's
say it's sitting here in this house right here,

it's broadcasting, it's like,
"hello, hello, hello, hello, hello,

I need an IP address [inaudible]" you know,

it's sending these broadcast messages
like I had to try and get one.

But if there's no DHCP server
out here, nothing ever comes back

and that computer goes, well, what do I use?

What it will do is it will just pick
something from that range, if it's a Windows--

and again, this feature can be turned
on and off, but it will just say,

well, I guess I'll be 169.254.50.22.

Now, it's smart enough to actually
send out a little beacon saying, hey,

is anyone else out there 169.254.20.52.

And, you know, if it doesn't [inaudible] okay,
that's the IP address that I'm going to use.

So, usually if you see this
address, something bad has happened

because you don't usually want to see those.

But what that allows you do is kind of
build your own little local area network.

You won't be able to get on the internet with
it, but your own little local area network

if things break down or you just
don't want to think about DHCP.

Loopback addressing.

When the inventors of TCP/IP thought
of it, they thought, you know,

we want to dedicate this massive
range of addresses for testing,

you know, for testing connectivity.

So, they took, you know, 4 million.

I mean just millions of addresses that--

actually 16 million addresses that are
in this range right here and they said,

we're going to use those for
testing, but really people ended

up using only one of them, 127.0.0.1.

If you go to any device in the world that
is on a network and you ping 127.0.0.1,

you'll always get a response back
as long as the device is actually,

you know, has a network card inside of it.

That is considered a loopback address.

It's an internal test that it uses to, you know,
just does a little cycle saying, okay, did--

you know, am I a valid network
adaptor, do I have TCP/IP enabled.

So 16 million addresses reserved
right there, and we use one.

And so that's the one that
we use for loopback testing.

The last one, last one I say special addresses,
and I mentioned this in the last nugget,

but I want to specifically bring it up here.

The first and last IP address of
a network or a subnet-- or is in--

or I shouldn't have said
subnet, it just came out.

The last or first IP address of a
network are the network and broadcast.

So if I were to say 172.30.100.0 or, you know,

that's the network we've been
using with this subnet mask.

I go, okay, well this is my
network and this is my host,

and I know that these IP addresses can
go zero to what's the biggest number?

255. Got it.

So, the very first address, the dot zero can't
be used because it represents the network.

It identifies it.

You see it inside of a routing
table of a router.

The last IP address also cannot be used.

I can't assign it to something because
that's considered the broadcast address.

So, if I wanted to send a broadcast
to everybody in that network,

I would send it to 172.30.100.255, and that
would reach everybody inside of that network.

Now, keep in mind it's not always a symbol.

I know it's easy when you're at this level.

It just be like, okay, anytime
I see zero or 255, that's bad.

I won't use those, I can't assign those.

Well, be careful.

I mean we're using this because
it's the most common subnet mask

in the world, but there are others.

Now let's say I did 172.30.0.0
with this subnet mask.

It's actually known as a class B subnet mask.

Now I'm saying this represents my
network and this represents my host.

By the way, that actually gives me 65,536
hosts that I can have in that range.

Now, if you just hard coded in your mind and you
go, okay, zero is bad and 255 is bad, well then,

you're going to have a lot of bad in
that range that aren't really bad.

There's only one bad or one that you
can't use, and that's 0.0 and 255.255.

This is the network and this is the broadcast.

Let's say there's only one
bad or only, you know,

two addresses, one of each that you can't use.

But also remember as you start counting, the way
this is going to increment is going to go 0, 1,

2, 3, 4, 5, 6, 7, 8 [inaudible] 20, 30, 40,
50, you know, all the way up to the point

where it gets to 255 then it's going to flip
this over and now we're at 1.0 and 1.1, 1.2,

1.3 [inaudible] 1.255 and 2.0, 2 dot--
and you see how it keeps going up.

So I could take the IP address
172.30.2.255 and I could assign that to a PC.

Whoa, weird man.

And if you're used to just seeing this and
going, oh, that's bad, or just seeing this, oh,

zero that's bad, then you're going
to miss a whole lot of IP addresses

in that range 'cause there's 3.0, there's
3.255, there's 4.0, and they're all valid

because remember in every single
network, there is only one network,

and there's only one broadcast address.

Everything in that squishy Oreo
center middle is totally useable.

Okay, last thing I want to talk about here.

Last gap that I want to fill in is the classes
of addresses and some of the terms like classful

and classless and kind of everything along
those lines, even some of the different kinds

of messages that we can have when we're
communicating using these IP addresses.

So, when the creators of TCP/IP invented
it, they came up with 3 standard classes

of addresses that exist, class A, B, and
C. And just by looking at an IP address,

I can throw one out there, 12.5.6.2.

Just by looking at that, I could go, oh,
that's a class A. I know that because you look

at that first digit right there, and if it's
between 1 and 127 it's considered a class A.

Or if that first digit is let's say it's
172.30.100.30, that's why [inaudible] kind

of been our mascot through this series.

I would go, "Oh, first digit is
172, that is a class B address."

So, it's real easy just by looking at
the first digit of the address to be able

to identify what kind of address it is.

But that's only half helpful
because it's like, well,

what does that do for me,
what are the difference?

Well, the difference is the default
subnet mask that these addresses have.

Class A addresses have the default subnet mask
255, 0, 0, 0 or a shorthand way to write that--

now, I will fully explain the
syntax later-- it's a slash 8.

Class B, the default subnet
mask is 255, 255, 0.0,

or a shorthand way of writing that is slash 16.

Class C is-- that's the most famous one, 255,
255, 255, 0 or a slash 24 is shorthand for that.

So, you know, rather than every single
time somebody putting, you know, 255, 2--

that a lot times they just go
slash 24 and people go, oh,

that's a class C address and what that is.

So-- or, you know, you're quickly
able to identify the subnet mask.

So, this was considered the classful
addressing of the old days, meaning, when--

if you applied this subnet mask,
these addresses, let's just grab one,

let's say 10.0.0.0 and it's classful
state has the subnet mask 255, 0, 0, 0.

What that means is I actually have one network,
it's the ten network, so everything that starts

with the number 10 is considered on my network,
and all of these octets are used for the host.

What that gives you is 16,777,214
useable host on that network.

It actually gives you 216 addresses,
but you remember the very first one

and the very last one can't be
used, so we have 214 useable hosts.

Now, initially like, "Oh
wow, that's a big network."

Well, it's just not, not
possible, it's not reasonable.

You can't have a network that big.

What that means is you've got a switch with, you
know-- or multiple, so I mean sure, you know,

give it the benefit of the doubt.

We can daisy chain multiple switches, right?

But you've got multiple switches
with 16,777,214 host, computers,

devices plugged into that one network.

There's no routers in between
them, doesn't work.

Why? Broadcast.

Broadcast will flood the network.

I mean Cisco, I would say just a guideline
and this-- I won't even say this is documented

but a kind of a rule of thumb, you
should never have a network above--

again, guideline-ish I would
say 500 devices at most.

And that's pushing it.

I mean I-- it's rare that I see a network more
than 200 devices that before we have some kind

of routed boundary in mind to kind
of keep the broadcast traffic down.

But I would say with modern processors
and all that, modern network technology,

you could probably get to 500 devices and
not see too much impact on performance.

But remember, each device that you add to that
network has its own broadcast that it's sending.

It's saying hello everybody, I
mean that your booting devices,

they're sending out traffic,
all that kind of stuff.

So, the more and more and more you add, the
more and more these devices have to receive

and process all of the traffic from all their
brothers that they don't even care about.

They're just dropping it.

They have to look at it.

They have to see what it is but they're
like, oh, it's not really for me.

I'm going to end up dropping that traffic.

So, a network of 500 devices is really
big, so you look at this and you go, okay,

we overshot that by 16 million some addresses.

So, that it's just not scalable.

So, what do they have in mind when
they created these monster networks?

Well, they had in mind that, you know,
well, we can use the classful version

and it will work, we can use those subnet mask.

They had in mind that people will use
something called classless addressing.

All classless means is that yes, yes, yes,
we have all of these classes of IP addresses

and we have the default, the
classful subnet mask that the standard

of TCP/IP said we should use
but I don't want to use that.

I want to put a class C subnet
mask on my class A network.

And now you can see why this one becomes
the most popular subnet mask in the world is

because what that does it say, okay,
well now I've got the 10.0.0 network,

there's my subnet mask, and
I've got one octet for host

which gives me 254 usable host addresses.

There's actually 256 that it gives you, zero
through 255, but you can't use two of them,

so 254 usable host address, and
that's a good reasonable amount.

So, what you see all over the place is people
applying class C subnet masks to all kinds

of different addresses inside of their
network, all kinds of different networks and--

but I mean what this does is give you
a ton more networks that you can have.

I can have the 10.0.0.0-- well, 10.0.0, 10.0.1,
10.0.2, I mean just a ton of different networks

with little packets of 200 some host.

You know this goes zero through
255, and each one of those networks

that I can use wherever I want to in my network.

I could start assigning those
to different offices.

So, I can have you know corporate up here which
is 10.0.0 and that has a connection over here

to our Arizona office and they use 10.0.1.

I mean, we start assigning these to
different networks all around our organization

and each one of those network support
254 useable IP addresses inside of them.

So, that's, that's the different
classes of addresses that exist

and at least what subnet mask they
start with, but I can change that.

Now, you're never supposed to go backwards.

You know, for example, I can-- I
could some-- you know, and this is--

this is actually by the way
considered subnetting

when you apply a different subnet
mask than what the standard allows.

But you're not supposed to go backwards.

I'm not supposed to say 192.168.0.0
with a class B subnet mask on it.

You know, you're supposed to
only go further, not pull back.

However, I've seen-- actually I
just started managing a network

where somebody used 192.0.0.0 you know with a--

they put a class A subnet mask on
this class C network and it worked.

I mean I had to go to therapy to handle
that network but at the same time it worked.

So, while you're not supposed to you-- you
know, in this classless world you can do it--

well, it's almost, you know,
two different meanings though.

In the classless world that we live in,
you can do almost anything that you want.

So, what about these other ones?

Class D and class E are untouchable.

Like we don't-- unless you have a very
specific reason like multicasting,

you don't use those in your organization.

So let me, let me fill in.

What are the different kinds
of communication that exist?

There's unicast, multicast, and broadcast.

Broadcast we know is one message to everybody.

Unicast we know is one message to one device.

It's, you know, the single device communication.

Multicast is the one in the middle.

It is one message to a group of devices.

This is by the way the future of communication.

We're going to see a lot more of this as time
goes forward but just to give you an example,

things that would be very useful for
this is like internet radio or imaging.

For example, let's take imaging.

Some of you may do this in your organization

where you have a server and
you have all these PCs.

Maybe you've got hundreds of PCs where
instead of going and installing, you know,

Windows on each one of those PCs by hand,
you create a master image and you send

that to all the PCs, so they-- you know
they all have the same look, the same feel.

You don't have to manually install Windows.

I mean imaging, like Acronis is
imaging, Norton Ghost is imaging,

these are applications that do it.

So, I set up a server which
sends an operating system

like an actual operating system image on there.

If I use broadcast to do that, it's
a mess because he send one message

to every computer just throttling
the network with all this,

this traffic because he's sending
a full blown operating system.

I mean it fills up a hard drive essentially
with the amount of data that he sends.

So it floods the whole network and our
whole network would probably go down until

that guy stopped broadcasting
'cause he's sending so much data.

If I did unicast, which is
valid, but each computer that--

that I add to the image stream
slows it down because he--

you know he starts off maybe
imaging at gigabit speed to this guy

and then I add another computer and now he has
to divide it because he's sending one stream

to him, one stream to him,
so each person gets 500.

And now we divide it again,
each person gets 333.

Divide it again, each person gets 250.

So it gets slower and slower and slower the
more devices you add to it at the same time.

Thus, multicast is born.

What multicast does is assign-- think
of multicast like a radio frequency,

like you tune into a radio frequency in the car.

What will happen is these guys get assigned
a radio frequency multicast address.

Let's just say 225.1.5.32.

I just made that up.

They all actually get assigned that same address
which, well, that's weird right there, right?

Because we're used to things
all having their own IP address.

But in multicasting, they all tune it-- I mean
don't think of it as being assigned an address.

They all tune into that frequency.

So, that the server starts
broadcasting [inaudible].

Sorry I can't say that.

The server starts sending to
that IP address or broadcasting

on that frequency is what I was trying to say.

But really it's just sending one stream to
that, that frequency, that address right there,

and all the computers tuned into it get it.

The ones that aren't tuned in don't.

So, they don't-- they don't get throttled
like a broadcast would and pummel the network.

So, multicast can be really
handy for doing things like that.

The final one is experimental, that's why I
brought our little scientist onto the scene.

You know, this was just a whole chunk of
addresses that when they create it they're

like we want this to like embed
in lab rats and things like that.

We definitely can't use or
touch those address ranges.

Whoo, a lot of gaps to fill in, right?

So, what did we see and more importantly, what
do I want you to do with this information?

What we saw was how a device gets an address.

Just what are static, dynamic,
you have multiple addresses,

even multi-homing connecting
it to two different networks.

We saw public and private addressing where
each would be used and what those ranges were,

along with the classes of addresses which
are classful versus classless networking,

even types of communication, unicast,
multicast, broadcast, I mean just a lot of--

and this is just kind of like a mega fact
nugget filling in a lot of that information.

So, what do you do with this?

Well, it depends on-- if you're studying for
the exam, I would highly suggest and that--

I won't even focus this for the exam.

If you want to be in networking, how about that?

I would highly suggest memorizing.

There's always going to be some
memorization, those private ranges.

Be able to see an address and be like I know
exactly what private address, if that's private

or public, and if you just see an address,
you should know what range it came from.

Thankfully, you can focus your
efforts on just class A, B, and C,

that just remember those ranges and where are
those read, along with what subnet mask each one

of those classes have and, you know, understand
that concept 'cause we'll see it again

of classful using the default subnet
mask versus classless IP addressing.

So, all of those just, you know, that's
brute memorization, get those things embedded

in your head using whatever method you like to.

Then, I'd like you to go on your device,
whether it's at home or in the office,

and pull open a command prompt, do an IP config.

Identify, you know, what network am I on?

What kind of class of network?

Am I using classful or classless addressing?

You know this, this I'm looking at is classless
because this is a class B address right here

with a class C subnet mask on it.

So, identify those kind of things.

Start thinking about what other IP
addresses would be in the same network as me.

You know, find out am I using DHCP?

You know, how do I know that?

I mean, I'm just looking and it come back.

Now just looking at that
that address, how do I know?

Well, you can do an IP config forward slash
all and that will tell you not only, you know,

if your DHCP which I can
find out looking up here,

but also what DHCP server
gave you that IP address.

In my network, I use the
router as the DHCP server.

I don't rely on a Windows
server where I come from.

I can even see, you know, some
information like what DNS servers I have.

Those servers resolved the name to-- name
to IP address, so when I type in Google,

I actually send a message to this
server saying, well, who is Google?

And the server comes back and
says, well, this is who Google is.

Start playing around with IP config.

I know, I know it's a Cisco world that we're
talking about here, but Cisco does require you

for the exam to know some of these commands like
IP config renew or release, so, I can say well,

I want to get rid of my IP address.

Whoops, we just lost network connection.

And so I, you know, and just like
that I kicked off my IP address,

lost all network connectivity.

I can come in there and say well,
you know let's do an IP config.

Again know that whoo, look at that.

See, look at all this action I'm having to do.

I mean do this, try this on your own device.

It generate, it says, well, if I'm not going

to use the DHCP address, then
I'm going to make up one.

Remember those 169.254 addresses.

I can say, yeah, well, you know, I'd
rather be connected to the network,

so I'll do an IP config forward slash
renew to get my-- and like please--

to get my IP address back
and now I can use it again.

So, again, play around with those.

Experiment.

See what you're using in your network
so that this becomes real to you.

I hope this has been informative and
I'd like to thank you for viewing.