CIDR, well, its pretty simple.
Take an IPv4 address of the form xxx
![Mad :x :x](/community/styles/default/xenforo/smilies/icon_mad.gif)
xx
![Mad :x :x](/community/styles/default/xenforo/smilies/icon_mad.gif)
xx:Xxx. There are 5 classes of IP addresses from A through E. And each address has two parts - a network part and a node/host part.
A Class A IP address' first bit is always 0 and the next 7 bits identify the network. The last 24 bits (the three dotted decimal numbers) are assigned to host addresses. Which translates to addresses between 1.0.0.0 to 126.0.0.0, leaving a total number of 16,744,214 unique hosts/nodes.
A Class B address has 10 as the first two bits. The first 16 bits identify the network and the last 16 bits identify the nodes, so this IP address is neatly split halfway between networks and nodes. IP addresses for this class start from 128.1.0.0 to 191.254.0.0 and can accomodate 65,534 hosts.
A Class C address has 110 as the first three bits, with the next 21 bits being the network address and the last octet is the node part. So we have IP addresses from 192.0.1.0 to 223.255.254.0 and can accomodate 254 addresses.
A Class D address has the first four bits as 1110 and is usually used for multicasting and has an address range from 224.0.0.0 to 239.255.255.254.
Class E is reserved for research and is not available for use on the Internet.
The problem comes when assigning IP address blocks to users that request them. If a company needs 300 IP addresses, they cannot go in for Class C, obviously and class B is too much, wasting 65,234 addresses. If they go in for two Class C addresses, they end up creating two domains within the same company.
To overcome this problem, the CIDR was developed. What it does simply is that enables you to use subnet masks to customize the length of the network and node addresses. So instead of the rigid 8/16/24 bits that the previous assignments used, CIDR allows you to use any bit assignment from 13 to 27 as a "prefix" on the IP address, specifying how many bits are used to specify the network address. A CIDR address is usually written as a standard IP address along with information on the prefix. So the address would usually be something like:
xxx.xxx.xxx.xxx/AB where AB is the CIDR "prefix"
So, a prefix of /27 would mean that 27 out of the 32 bits are used for the network address and the remaining 4 bits are used for the hosts, allowing you to use that address for 32 hosts. So, what usually happens is that your ISP is allotted a CIDR block depending on their size, so (just an example) while VSNL would be allotted a /13 block prefix, they would in turn allot /15 or /18 blocks to ISPs like Sify, etc. who in turn allot it to resellers, who in turn allot it to you, etc. Its not the exact way its done, I was just giving you an example.
So, you would use your CIDR block to calculate addresses for your organization. So, if you are allotted an address like, say, 10.0.0.12, and you wanted to run 1000 computers on that IP addressing scheme, you would pick the CIDR prefix of /21 that allows you 2048 hosts, then you would have a subnet of 255.255.248.0 and a total number of 8192 subnets and 2048 nodes per network. For a great subnet calculator, visit this page here:
*www.subnetmask.info/
Hope that clears things up a little.