A Couple of simple TCP/IP Questions... 7

[tfccom]

Which is better... a 10.0.x.x based network, or a 192.168.x.x network, and what is the difference?

What's the difference between a class "B" and class "C" network?

Trevor Farren

 

[jimbopalmer]

Class B the 10.0.x.x network allows you to accumulate 64 thousand ethernet devices. you would then build an internal infrastructure to sub-divide the thousands of devices.

if you only need 254 devices a Class C network 192.168.x.x is plenty and simpler as it is smaller. most consumer Routers, as an example, are pre configured for a Class C net. I tried to remain child-like, all I acheived was childish.

 

[Guest_imported]

The way I see it there really isn't any difference other than the number of hosts that are possible. But I don't think you'll need more than a few hosts if this is a home network. Both are private network schemes (not routable on the Internet). The question really is how many hosts are you planning. If less than 254 then take your pick.

 

[johnny99]

Don't think so much about numbers.
What's importend is to have an IP-number plan.
Class A, B, C is a thing of the past. It has no real meaning today. Today a class A address space is called /8, Class B /16 and Class C /24.
Your networks can be /28 if that's the right size for you.

Two things that always give people problems are that they made a wrong plan in the beginning and the day they have to connect there network to other peoples network.

I think the main reason why people still use the old class A, B and C is because they can't calculate the netmask for eg. a /28 or /20

Some usefull links:

http://www.ripe.net/ripencc/pub-services/db/rfc1918.html

http://www.faqs.org/rfcs/rfc1918.html

http://www.faqs.org/rfcs/rfc1519.html

http://public.pacbell.net/dedicated/cidr.html

/johnny

 

[ChrisAC]

LOL. You're right about that! You wouldn't believe the number of IT managers I speak to who can't work out subnet masks and don't know what /28 or /29 means!! This is not all of them by any means but I've spoken to a few.

Strictly speaking the mask doesn't determine the class. Before the use of CIDR all subnet were either /8, /16 or /24 for class A, B or C networks (ignoring the others!!). However people assume that the subnet mask determines the class, when it was in fact the value of the first octet.

For example, the RFC's stated that for a class A network, the first bit of the first byte will always be 0. So, a class A address will always have the first octet between 0 and 127, although we all know that 0 isn't used and 127 is a loopback address. So, in theory you have,

00000000 = 0
01111111 = 127

A class B always has the first two bits of the first octet as 10, giving

10000000 = 128
10111111 = 191

A class C starts with 110, giving

11000000 = 192
11011111 = 223

So, a 10.0.0.0 /16 isn't really a class B network, but a subnetted class A network.

However, having said all of that johnnny99 is spot on when he says that we don't really work with classes anymore as most ranges are "classless". When a customer comes to us with a request for an internet connection they have to fill out a RIPE application form to justify the number of IP's they require. This will normally be /30, /29, /28 or /27. Under normal circumstances these ranges will provide enough IP's for what they require which is normally one for the router, one for the firewall or proxy and then mail servers, web servers etc.. What we do get though is the odd customer that has less that one hundred clients, one firewall and a mail server and still applies for a /26 or something stupid like that.

Anyway, great links johnny99!!

Chris.


************************
Chris Andrew, CCNA
chrisac@gmx.co.uk
************************

 

[GlenJohnson]

If you want some more good (and free) information, go to

http://www.learnntcpip.com.

Good luck. Glen A. Johnson
Microsoft Certified Professional

"Things are not always as they seem;
the first appearance deceives many."
Phaedrus (15BC - 50AD) Roman poet, short-story writer