/4 subnetting

oxzgan

hy guys,

basically i only know the classless networks starting from /8 i mean 255.0.0.0
and to /32 ie 255.255.255.255

i found some books saying about /4
is there classless networks below /8 ..ie /7 , /6 , /4 etc

WillTech105

Classless networks under /8? I guess there CAN be -- why you would would be a different story.

The TCP/IP Guide - IP "Classful" Addressing Network and Host Identification and Address Ranges

According to this there'd be at least 2 for the network ID for it to be considered a Class A Addr. Hope this helps!

Qord

A /4 would be used for point-to-point links where all you'd need is two addresses. Look more into VLSM.

Forsaken_GA

Qord wrote: »

A /4 would be used for point-to-point links where all you'd need is two addresses. Look more into VLSM.

No, that would be a /30 (and depending on your IOS, maybe a /31)

A /4 would have 268,435,456 million IP's available for hosts. You'd basically be aggregating several Class A's, which is a bad idea, since most Class A allocations tended to go heavily underutlized

phoeneous

Qord wrote: »

A /4 would be used for point-to-point links where all you'd need is two addresses. Look more into VLSM.

You might want to triple check that.

/30 is your friend.

jason_lunde

phoeneous wrote: »

You might want to triple check that.

/30 is your friend.

RFC 3021... /31 is even more your friend

Edit...nevermind didnt see foresaken's post above...now

oxzgan

i found some books using /4 ip address .
I am only familiar with /8 ( class A ) / 16 class b /24 class and VLSM too
.
Is it possible to have them below 8 like /7, /6, /5, /4 etc?
I posted this question on ccna section ,but doubt remains same.

as far as study
i found this .
/8 is class A 1-126 range ( binary 0xxx xxxx to 0111 1110 )
/16 class b 128 - 191 range ( binary 10xx xxxx to 1011 1111)
/24 class c 192-223 range ( binary 110x xxxx to 1101 1111 )

for class A we all know that 8 bit network portion and 24 bit host part . range indicated only the first bit should remain 0 and other bits can be anything.
is it because /4 address used . like first 4 bits remain same?
am i correct?

tiersten

oxzgan wrote: »

Is it possible to have them below 8 like /7, /6, /5, /4 etc?

You can do it if you really wanted to. There isn't any good reason to so that I can think of. It'd have to be an internal private network anyway as you can't get an empty /4 out on the real IPv4 internet. If your internal private network needs this configuration then its incredibly badly designed.

oxzgan wrote: »

I posted this question on ccna section ,but doubt remains same.

You don't need to and shouldn't post the same question in multiple forums. It just means you get duplicated replies and not everybody may be reading the other forums you're posting in i.e. CCIE.

Forsaken_GA

oxzgan wrote: »

i found some books using /4 ip address .
I am only familiar with /8 ( class A ) / 16 class b /24 class and VLSM too
.
Is it possible to have them below 8 like /7, /6, /5, /4 etc?
I posted this question on ccna section ,but doubt remains same.

as far as study
i found this .
/8 is class A 1-126 range ( binary 0xxx xxxx to 0111 1110 )
/16 class b 128 - 191 range ( binary 10xx xxxx to 1011 1111)
/24 class c 192-223 range ( binary 110x xxxx to 1101 1111 )

for class A we all know that 8 bit network portion and 24 bit host part . range indicated only the first bit should remain 0 and other bits can be anything.
is it because /4 address used . like first 4 bits remain same?
am i correct?

Is it possible? Yes. And it's not even theoretical, IOS will let you configure it.

stark(config)#int lo0
stark(config-if)#ip add
stark(config-if)#ip address 10.0.0.0 240.0.0.0
stark(config-if)#do sh ip int lo0
Loopback0 is up, line protocol is up
Internet address is 10.0.0.0/4
Broadcast address is 255.255.255.255

CIDR removes the concept of classful addressing entirely (that's why the C stands for classless). With CIDR, you can notate any block of IP's in any way you want from 1 to 32 bits.

Practically speaking, any network engineer that implements a /4 in production better have a damn good reason for doing so, or else he's joined the ranks of bad designers

Also note that I supplied the IP as 10.0.0.0. This is valid, since we're talking about a /4. A /4 mask covering the 10 network would technically cover 0.0.0.0 through 15.255.255.255. This would be one of the few cases where subnet zero really is still unusable for host addressing.

Basically a /4 aggregates 16 /8's. That is one big ass broadcast domain

notgoing2fail

Interesting question. I've never even thought about a \4.

I agree with others, private LAB use only....

oxzgan

Forsaken_GA wrote: »

Is it possible? Yes. And it's not even theoretical, IOS will let you configure it.

stark(config)#int lo0
stark(config-if)#ip add
stark(config-if)#ip address 10.0.0.0 240.0.0.0
stark(config-if)#do sh ip int lo0
Loopback0 is up, line protocol is up
Internet address is 10.0.0.0/4
Broadcast address is 255.255.255.255



Basically a /4 aggregates 16 /8's. That is one big ass broadcast domain

Thanks for detail reply.But one doubt how you can understand that 240.0.0.0 is subnet mask for /4

jason_lunde

oxzgan wrote: »

thanks for detail reply.but one doubt how you can understand that 240.0.0.0 is subnet mask for /4

128 64 32 16 8 4 2 1
1 1 1 1 0 0 0 0

128+64+32+16=240

chmorin

jason_lunde wrote: »

128 64 32 16 8 4 2 1
1 1 1 1 0 0 0 0

128+64+32+16=240

I'm curious how he got his CCNA without understanding the details in subnetting. (That may have been mean, sorry.)

There are three classes to subnetted networks: A (/8 ) B (/16) and C(/24). [There are others actually which include some excluded addresses spaces and multicast].

With modern networking, you are only limited to binary math with four octets in dividing network and host bits in an IPV4 address scheme. Which means with slash-dot notation you can come up with any devision of networks you want. While /4 for is ridiculous, it is possible. So are /3 and /2, and I guess even /1.

tha_dub

chmorin wrote: »

I'm curious how he got his CCNA without understanding the details in subnetting. (That may have been mean, sorry.)

We've all done things in haste without thinking them through carefully.... It's best to do here rather than in a production environment I once replaced a bad dual channel PRI card in a Nortel cs1000 during production with the same bad card and then proceeded to troubleshoot why it wouldn't come up.... Thankfully it didn't take me long to figure out what I'd done but I had engaged tier 3 support at that point to whom I had to explain my mistake...

Qord

tha_dub wrote: »

We've all done things in haste without thinking them through carefully....

I'll vouch for that.

Some of us(me) need to take a little more time before we post.

tiersten

Qord wrote: »

Some of us(me) need to take a little more time before we post.

Eh. Better to ask random questions on a forum than to not ask and then actually break something on a real network because you didn't understand.

phoeneous

tiersten wrote: »

Eh. Better to ask random questions on a forum than to not ask and then actually break something on a real network because you didn't understand.

Agreed. Unless your coworkers or boss also post on this forum and they know your username

chmorin

tiersten wrote: »

Eh. Better to ask random questions on a forum than to not ask and then actually break something on a real network because you didn't understand.

Haha Alright you all win, better off post here and understand properly. I for sure have asked stupid questions on here as well, only to find our 30 minutes later how stupid I actually was. But if someone with a CCNA walked up to me and didn't know binary to save their life, I would have a few questions to ask is all... not that this is the case here, but you get my point.

I once spent two days asking questions and trying to figure out why my config on my CME router was not working only to realize I had not created my CNF files... and there are tons of stupider things on the list as well. Probably half of my posts are questions.