All language subtitles for 23 - Routing - Creating Subnets Based on Host Requirements-eng

>> It's our second flavor of subnetting now.

Subnets based on host requirements
rather than network requirements

as we did in the previous nugget.

So, just like the previous nugget we're jumping
straight in the water with four unique examples.

Now, this first example looks
the same and feels the same

as the first example we saw in the last nugget.

However, there's a major twist
to it that makes it different.

Notice, this organization has purchased
the Class C address 216.21.5.0.

So, immediately I think slash 24.

That's the Class C subnet app.

And we'd like to create networks
of thirty hosts each.

Ding ding ding ding.

That's the unique twist.

So, in the previous nugget, we said,
how many networks do we want for this?

And we counted them.

OK. One network, two networks,
three networks, four and five.

And so we figured this question
based on five networks and kind

of broke up the subnets within each one.

Now, this one we've stated the
question a little differently.

We said, well, honestly I'm not, I don't
really, my main focus, let me put it that way,

my main focus is not on how
many networks I can get.

My main focus is that I need networks of thirty
hosts each because my office is, you know,

that office has twenty people, that
office has twenty eight people,

that one has fifteen people, so I really
want to create this so it fits my offices.

So, the offices will have 30 usable IP
addresses that they can use per office.

The good news is that the process is almost
exactly the same as it was previously.

So, let me quantify almost.

So, first I've written up
our number again up here.

216.21.5.0 is the network and
we wanted 30 hosts per network.

Right? So, 30 hosts per network.

So, step one is determine the number
of hosts and convert to binary.

So, we started with 30 hosts and
we'll write that as a binary number.

Now this is the same as it was previously,

it's just it said convert the
number of networks to binary 4 2 1.

So, we'll do thirty, so that'll be 0 0 0 1, so
30 minus 16 and I know, there's the shortcut,

but we're going to do it the
long way to start off with.

4, 14 so 1 minus equals 8,
so that'd be six, right?

1 1 0, so the binary equivalent
of 30 is 00011110, good.

OK. Step 2.

Reserve bits in the subnet
mask and find your increment.

Now, we're back to that Class C network,
255.255.255.255.0 So, I'll go, 1,2,3,4,5,6,7,8.

1,2,3,4,5,6,7,8.

1,2,3,4,5,6,7,8.

1,2,3,4,5,6,7,8.

Good. Now, here's where the twist comes
in, here's where the almost comes in.

Now, previously, I kept asking the
question, OK, are we after networks

or are we after hosts per network?

And the answer in the previous
nugget was always networks.

We're after networks, networks, networks.

We kept emphasizing we're after networks.

And network bits were the ones, right?

We were saying ok well then lets put ones
here and lets stretch it out this way.

Now that all changes now,
because instead of saying I'm

after networks, I'm after hosts per network.

Now, keep in mind, I'm still subnetting,
it's not like I'm moving those ones backwards

or anything like that, I'm still subnetting.

However, my focus, my main focus
is not how many networks I get,

it's how many hosts per network I get.

So, what I'm going to be doing is
saying, instead of looking for, you know,

and we've got to identify, it takes five
bits to get the number thirty, right?

Just like we saw in the previous nugget, so
instead of saying I need five network bits,

because that would be the previous style,

is saying I've got 30 networks
that I'm after, no no no no.

I'm saying, I can't get the number
30 with any less than five bits,

so I need to save five host bits.

1, 2, 3, 4, 5.

Did you notice what I did there?

Two things: one is save the five host bits.

But, also did you notice
the direction I was going?

Instead of going left to right, like I did
with the networks, I went right to left.

Because that's where my zeros exist.

That's where the client bits are at.

So, I'm saying, as long as I have five
bits, I know that I can get the number 30.

OK? All the other ones I
don't need for the clients.

I've met my requirements.

I can get my 30 hosts with my five bits.

So, the rest of them can go to the
network site because guess what?

I'm still subnetting, right?

This is still subnetting, I'm still
breaking one network which was 216.21.5,

into multiple subnetworks, but I just
wanted to make sure that those subnetworks,

however many I get, will have at
least 30 hosts per subnetwork, right?

OK. Everything else is the same.

That's the story, is my focus is on the zeros,
so now let's figure the rest of this out.

It says, so we've changed our subnet mask, we
know what it's going to be; it's 255.255.255

and then this is the lest octet,
so it'd be 128 plus 64 plus 32.

That'd be 224, or if we're using citer notation,
we just moved from a slash 24 down to a,

we added 3 more, so that'll be a slash 27.

Again, just, these are two ways
of writing the subnet mask.

That's it.

No other concepts than that.

Now, the increment is still the
lowest network bit converted back

to a decimal number; that would be 32.

Right? Lowest network bit is the decimal, and
by the way, when we're figuring out the number

of hosts, if this number is ever
less than your host, be concerned,

because that really defines the ranges.

How big are they?

So now, I can come down here to
figure out my network ranges.

216.21.5.0 is where we begin, and then we just
start adding our increment of the third octet.

You see how everything else is the same?

216.21.5.32 216.21.5.64 and
down and down we go: 96, 128,

and this would continue on down off the screen.

Fill in the end ranges.

This goes to 216.21.5.

what's the last IP address?

31, right?

216 also realizing that we can't
use that broadcast address because,

well because the last idea,
because it's a broadcast,

so usable, we actually have 1 through 30.

We have exactly 30 usable hosts per subnet.

Just one of those weird values that
happened to land on the exact amount,

but we would just continue this trend all
the way down, filling in the network ranges.

Awesome. So, again, the slight tweak that
we get is right here in the second step,

to where we're no longer
focused primarily on adding ones.

We're focused on saving zeros.

I was teaching a class one time, and we
were trying to figure out if this was long,

like when I had first created the
subnetting method, and we were trying to,

and somebody was like, how do I remember that?

You know, how do I remember when I'm going right
to left or left to right and ones and zeros

and I was like, I don't know, how can
we remember, guys, to save the host bits

and this guy from the back, he was totally
random, he yells out, save the hosts!

And we all kind of looked
at him, we were like, what?

And he goes, well it's kind of like, you know
save the whales, you know, like save, you know,

we have to save the hosts, so that
became like the theme from there on out,

is any time we saw one of these
examples, it was like, ok,

are we trying to add more
networks or save the hosts?

And everyone would yell out, save the hosts!

So, I have been teaching in classrooms
for a long time the subnet method,

and it was funny because my wife, my wife
actualy went to one of my CCNA classes early on,

so I can say she's a CCNA sort of,
just cause she attended the class.

And, so I took her with me to Cisco networkers;
this was maybe 5 years ago, it was before kids,

bc, when we could do that, and so we were
there and she was at this big event with me,

and we were, just a ton of people just crowds
and crowds of network people at networkers

and there was a guy who from across
the room goes, save the hosts!

And it was funny, my wife looks at me and
she goes, someone did not just say that?

And I was like, where is he?

My brother, you know and so it was
funny, we searched and I was like hey!

And it was just a guy that had been
in one of my classes so anyway,

this save the hosts thing has been around
for a long time so I'm going to use it.

I'm going to keep saying that when
we're going through these examples.

OK. Example number 2.

We've got a Class C address again,
same story, 195.5.20.0 and we're going

to need 50 hosts per network,
so a little larger organization

of 50 hosts per network is
what we're after, so let's go.

I've got 50 is my magic number.

I'm looking for how many
bits to get the number 50?

64, 32, 16, 8, 4, 2, 1.

So, I'm going to go no, no, yes.

Now, I'm going to start invoking shortcut
number one right here, I'm going to say ok,

I can stop because now I know it gets, excuse
me, it takes six bits to get the number 50.

That's an easy one, right?

That's 48, 50, but nonetheless, it still
takes 50 bits to get to 150, so now I look,

it's a Class C, so let's
do this, I'll do slash 24.

And again I'll start showing you
now we've seen a few of these.

Let me give you some shortcuts that I do
when I'm writing it on paper, like on a test

or something like that, what
I'll do is I'll say ok, slash 24.

I know that's 24 ones.

I know that's the first three octets,

so when I'm used to saying it,
I'll go 111....111.00000000.

I always write the zeros.

Zero. Because that's our playground,
that's where we get to work.

So, essentially, again, if this
messes you up, don't do it.

But that's what I do, just so I
don't have to write all three octets.

This is, this represents 24 bits.

Right? 24 ones split between the
first three octets of the subnet mask.

So, what I'm focused on is the zeros.

Now, again, back to my scenario.

Are we creating 50 networks or are
we saving 50 hosts per network?

Save the hosts we all yell, hands in the air.

So, we're going to focus
instead on saving six bits.

1,2,3,4,5,6 must remain zero because I can't get
my fifty hosts with any less than that and then,

shup shupe, those two can flip over to
one, because I'm still doing my subnetting.

So, now I've got my new subnet mask.

It moves to a slash 26, or if I wanted to write
that in decimal it would be 255.255.255.192.

That'd be 128 plus 64, right?

So, the lowest network bit, converted
back to a decimal is 64 in this case,

which makes me feel warm and fuzzy because
I'm looking, ok, I needed 50, I got 64, right?

So, I'm going to come down here
and let's use that increment, 195.

Oh and be so careful.

I've seen on Cisco exam, I don't know
why Cisco does this, but, for instance,

everybody's used to a private IP address,
like 192.168.1.0 and all those kind of things.

Cisco will have a simulation on
their exam where they're like, ok,

go ahead and assign the IP address
192.186.1.9 to that interface.

I don't know why they do that, because it's
so, like you see that and you're so used

to seeing this that you automatically
transpose them when you type

and then you assign the wrong
address, but maybe they're just like,

we want to make sure you're
really reading the question.

But nonetheless, 195.5.20.0
is where we're beginning.

It's in the fourth octet, so
I'm going to do 195.5.20.64.

I'm going to take this one all the way.

195.5.20.128.

.192. Now, what happens if we add 64 to 192?

We get 256.

Again, just like the last nugget, don't panic,

it's ok that you missed it,
that means you're done.

So, I'm going to fill in the in-ranges.

This goes to 63.

127. 191. 255.

So, what does this tell me?

This tells me that when I'm getting
networks of at least 50 hosts per network,

I can only get 4 networks
out of a Class C block.

I just simply run out of IP addresses if I
try to go more than that, so that's good,

it meets my requirement and it also gives
us a chance to go back to that formula.

Remember that?

Two to the power of bits.

So, I'll say, if I'm curious about the
networks, I'll do two to the power of two

because there's two network
bits, so that's two times two

or remember these guys to the power of zero.

1,2,3,4,5,6.

So, I can say two to the
power of two is actually four.

So, I get four networks and that lines up.

Or, if I'm saying, well how
many hosts do I really get?

I can do two to the power of,
well we got six host bits.

So, look up here, 6.

So, that'd be 64.

But, I always have to subtract two,
so I actually get 62 usable hosts,

because I've got to account for the
broadcasting network per network.

Yes, it's more than 50 but that's
just how the binary works out.

There's no binary increment, you know
that, the next one down would be 32.

Do you kind of see how these numbers just seem

to repeat themselves again
and again we see them.

And that's how you can get so comfortable with
networking and so comfortable with subnetting,

is once you get used to seeing those eight
values, you just kind of see them used

and popping up in different places.

Alright, third example.

Let's go beyond the class c. We've got
150.5.0.0 and we want 500 hosts per network.

That's the biggest one that I
would suggest you would create.

Otherwise broadcasts get a little crazy.

So, let's do it.

We've got 500.

How many bits to get the number 500?

Well, let's write it up,
128, 64, 32, 16, 8 4, 2,

1 and immediately realize
these as I'm writing this.

We need more digits, we need more binary
values because we can only get 255 with 8 bits.

So we multiply by 2.

We go, ok, 256, 512, 1020, sorry
I'm going a little bit far,

but the point is, we go OK, there.

That's where our first one goes
because I can't subtract 512 from 500,

so if my first one goes right there, again I
could find the whole binary value, but instead,

I look and I go, well to get
the number 500, it takes 9 bits.

I write my original subnet mask which is
slash 16, and again I could use the shortcut,

but I'm going to write it all out this time.

1,2,3,4,5,6,7,8.

1,2,3,4,5,6,7,8.

And 1,2,3,4,5,6,7,8.

I find just when we go to Class B, Class C, I
just like seeing the whole thing right there.

Ok, so, now again, I'm asking the
question, am I creating more networks?

Meaning focus on one, or am I saving the hosts?

Save the hosts, we all yell.

We've got to save 9 host bits, because we can't
get the number 9 with any less than nine bits.

It's like, we can't get the number
500 with any less than 9 bits,

so I'm going to go right
to left, 1,2,3,4,5,6,7,8.9.

Those 9 stay on the host side.

The rest of them, chup, chup, chup, chup.

Is that seven?

Seven. Flip over to the network side.

So all my new subnet masks becomes
a slash 23, which is 16 plus more.

Or, if I was writing in decimal, 255.255.254.0.

That there, those two are just
different forms of the same subnet mask.

My increment is going to be
that guy, who is a q, right?

Yes, he's in the same octet, but remember
each octet restarts the same increment,

so he is a two.

That's fine.

So, now I can use that increment
to find my network ranges.

150.5.0.0.

We go .5.2.0 .5.4.0 .5.6.0 and
down and down and down we go.

Right? Now again, same trick as the last nugget.

We don't want to take this through 1.0.

We want to go 150.5.1.255.

150.5.3.255.

Right? Last IP address before
4.0 and down and down we go.

Now, interesting enough, you might be asked,

well how many hosts per network
does that really give you?

Well, we look at our formula, we say, well
that would be two to the power of nine

because I've got nine host bits minus
two, because we've got to account

for the first and last IP address.

So, I'll just look at this and go,
ok this is two to the power of zero.

1,2,3,4,5,6, 7, 8, 9.

Nine. 512 minus two equals 510 host
bits, or hosts that I get per network.

And it's almost a perfect match
for what I was looking for.

OK. Last example.

We've got the ten network, Class
A network that we want to break

into subnets of 100 hosts per network.

And that's a big network to divide
up into all these little ones.

So, first thing to do is take
that 100 hosts, the hosts- Oh.

A little copy paste madness on
Jeremy, got caught up to him.

This should be hosts.

Hopefully the previous ones weren't like that.

If so, I'm sure you would have seen
a little bubble because I'll go back

and add a bubble there saying,
I really meant hosts.

So, determine the number of
hosts and convert that to binary.

So, I've got 100 hosts that I'm
going to convert to a binary number.

128, 64, 32, 16, 8, 4, 2, 1.

And by the way, the more and more you do
of these, the more you'll get to the point

where you just kind of look at them.

You kind of, I mean, again,
remember, it's eight numbers.

It can only be one of eight numbers, so
when you see 100, you may get to a point

where you don't even have
to write that up anymore.

You just look at it and go,
OK, well that takes 7 bits.

I know that 128 is the eighths bit, so it's
going to take seven bits to do that, so again,

you'll just get faster and faster
and faster at these, as you go.

Just to give you a good guestimate for doing it.

I would suggest before you take the CCNA
exam, that you should be able to work

through one of these guys in five minutes.

I know, I know that's aggressive, but five
minutes I think should be the maximum amount

of time that you spend on doing
an entire question like this.

That's kind of the speed
level that you need to be at.

Otherwise you might run short on your CCNA exam.

So, we've got seven bits that we're getting.

Reserve bits in the mask, so
I'm going to say a slash eight,

that's a Class A, is 1,2,3,4,5,6,7,8 dot.

And then you want to write out all the zeros.

1,2,3,4,5,6,7,8.

1,2,3,4,5,6,7,8 because you've got to be able
to see where these lines are being drawn.

OK, so are we creating 100 networks?

Is that type 1?

Or are we creating 100 hosts per network.

We're at hosts per network, so
save the hosts, save the host bits.

We're going 1,2,3,4,5,6,7, and then 1.

He flips over to the network side.

Now, whoa, what about all the rest of them?

They all flip.

That's my very fast writing.

They all flip over to the network
side because we graduate ourselves

from a slash eight all the way down to a
slash 25, as in all these become subnets.

So, what we get is we get fairly small networks,

100 hosts per network, and
a whole bunch of them.

You know, a whole bunch of these bits.

We say, well good grief, that would be, what?

Two to the power of, we added 8, 16, 17.

17 subnet bits to figure
out the number of subnets.

Are we going to figure that out with- First off,
they're not going to ask you that on the exam.

Cisco is not cruel, often, to
ask you a question like that.

What they'll do is they will,
they'll ask you something

that you can figure out just by multiplication.

What I would do if I got this,
is let's go to scientific.

Yeah, there we go.

So, we can actually do two to the power,
that's x carrot y, 17 equals 131,072 subnets.

So, what this little scheme gives us.

So, we've got a lot of networks
that we can grow into.

So, we go, ok, that's good,
we've got our subnet mask

if we want the decimal version
it's 255.255.255 dot.

what's that last one?

128. That's a new one.

We haven't seen that one yet.

So, those are two different
ways of writing our subnet mask.

Our increment happens to be 128 as well.

Same value as the subnet mask.

It's one of those rare circumstances.

So, now I can drop down and
figure out my incremements.

10.0.0.0 is what I was given,
that's where I began.

My increment is in that last octet, right?

So it's going to go 10.0.0.128.

10.0.0 dot.

Now what if we add 128 to 128?

We get 256, right?

No, we can't have that.

So, this goes back under
one of those rare rules.

I shouldn't say rare.

One of those rules to where,
let me fill in the n ranges.

This goes to 127 and if this was 256
down here, this would go to the 255.

So I just have to say, OK well
the next one then starts at 1.0.

1.128 and then 2.0 and then 2.128.

And it just keeps on going down and down
and fill in the n ranges, this goes to 127,

this goes to 255, do you see the trend?

So, I needed 100 hosts per network and I
ended up getting about 126 hosts per network.

So, our, you know, essentially two to the
power of seven minus two gives us the number

of hosts per network, and that
will give us our network ranges.

Now of course I have homework that I have the
written answers to, where you can download them

from CBT nuggets and get the answers to,
but let me just make a broad statement.

This is what I would like you to do as homework.

Let me make a broad statement.

At this point, I would hope that you're
feeling pretty, pretty warm and fuzzy

about the two different styles of
subnetting that we've talked about so far.

And I say that because, really, it's one style.

It's kind of the same three steps.

I'm just slightly tweaking it as we're getting
into this nugget, and just giving you a number

of examples that we've walked through.

If you're not feeling warm and fuzzy, to
where you're kind of like, it's still,

I don't quite get it, I would seriously
suggest running through those nuggets again.

Maybe just bounce it off somebody.

Have somebody watch it with you.

I will find that sometimes just talking
through my questions with somebody,

that just when you get to a point, you
know, even if the person has no idea.

If you're married this is the best thing for
your spouse to do, if they could help you out.

Just have them sitting next to you
and be like, ok, let me stop it,

when Jeremy said that, what did you understand?

Did you get that?

Because I will tell you, sometimes
even somebody that has no clue

about networking will be your biggest
ally, to just talk it out loud with them.

I will tell you, some of the most whacked
out network situations I've been in,

I'm sitting here, you know, as, when I talk to
your right now, is how I am most of the time,

you know, I'm just always, you know, I'm
thinking out loud, I'm pacing around rooms,

I'm going, OK, OK so if I got
this, and I'm like, I don't get it.

Why is this not, it should be working and you
know, God sends her in I believe, you know,

my wife always walks in the room at just the
right time, and she's like, what's going on

and I'm like, ah, it's just in
networking, don't worry about it.

And she's like, no, tell me.

And I'm like, ah, cause I know she
doesn't know what I'm talking about.

So I'm like, OK, I've got a computer over
here, it's not talking to the server,

I don't know why, and she's
like, uh huh, uh huh,

and she might just be like,
so, can he surf the internet?

And I'm like, well yeah, I
don't, that's, ok, hey go away.

You know, I've got it, hang on, wait a second.

So, a lot of times, she'll just
talking it through with somebody

who may not even know a thing about
networking can sometimes help you out,

so I would really encourage you to do that,
and then chalk through these four problems.

So, two styles of subnetting you now have down.

We've got subnetting based on the number
of networks, and then subnetting based

on the number of hosts per network.

Two more to come.

That would be reverse engineering a
subnet problem, and then the ultimate,

the pinnacle variable net subnet masking.

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