First and foremost, you should consult RFC1878. Use this if that is too advanced for you.
This is the first in my series of Internet engineering and network policy documents. It was originally thrown together as a reference guide for my co-workers and employees at UltraNet. It wound up being used as a training tool since it had the table as well as explainations. Not to toot my own horn, but I have yet to see another document, on paper or online, that transfers the clues as well as my own. But I AM biased. Very little of the original text was modified, except to modernize and remove company-specific parts.
If the terms here are not your speed, or this makes your head spin in some other respect, you should search for a simpler guide. (former link that was here is now dead)
The following table shows the relationship between IP network's block sizes, classful interpretation, netmasks, bit-length, CIDR "supernets", "cisco wild bits" and the actual bit-sizes of the networks.
An IP address on its own does not communicate what size network block to which it belongs. A netmask associated with an IP address describes "masked" portions [the left-most bits 'blocked out' by the ones in the netmask] of an IP address. These masked portions create the "network part" of the address, commonly called the "prefix".
The block size tells you how many addresses are within the network. The "bottom" [all-zero] is reserved to refer to the network itself and "top" [all-ones] is reserved as the broadcast address, so these are unusable. therefore, any given netblock has two less usable addresses that the literal block size.
Eg, a network 10.0.0.64 with a block size of four has only two usable addresses, 10.0.0.65 and 10.0.0.66.
The netmask indicates the octets that should be masked against the address to configure equipment or define routes with the apropriately-sized network. All leading [leftmost] octets in a netmask are ones.
Eg, a network 10.0.16.0 with a netmask of 255.255.252.0 would result in a network netmask pair of 10.0.16.0 255.255.252.0, while the previous example from block size would result in 10.0.0.64 255.255.255.252.
The cisco wild bits indicate the appropriate value for the octets of what is needed when creating a Cisco router's access list for a certain network size. For small networks, the final octet is one less than the the block size, merely because given the network number [the "zero" or "bottom" of the network], the "wild bits" define all the remaining possible values within the network. It is alwyas the inverse of the netmask.
The CIDR supernet notation indicates the literal bitsize of a network. "Supernet" is actually a misnomer, stemming from the perceived need to differentiate prefixes longer than 24-bits [subnets of "Class C" blocks] from those shorter than 24-bits in traditional "Class C space" (eg, the 207.* network), calling ther latter "supernets". The number actually represents the "prefix length" - what part of an address is the network-part, or tht which is "masked out".
Given that IPv4 addresses are 32-bits long, this is the most common and easy way of describing a network. A network of one device (a host route) is defined as easily as a network of two hundred fifty-six devices: 10.0.99.99/32 is a host and 10.0.98.0/24 is a "C-sized block".
This notation is the modern standard for describing network sizes.
The raw bits define the bits "within" the network. These are the inverse of the CIDR notation, and are a bitwise representation of the blocksize. Lastly, this is the exponent for "2" when describing the network mathematically.
The classful row merely gives a "backwards compatible" reference. All classful interpretations [Class A, B, C, D and E networks] of IPv4 space are archaic. Keep in mind that CIDR stands for "ClassLESS Inter-Domain Routing"; this row is provided to give a reference point when speaking to people that do not yet have a firm grasp of CIDR notation.
block size | netmask | Cisco wild bits | CIDR | raw bits | classful | common use |
---|---|---|---|---|---|---|
1 | 255.255.255.255 | 0.0.0.0 | /32 | 0 | Class C: subnet Class B: subnet Class A: subnet |
Host route |
2 | 255.255.255.254 | 0.0.0.1 | /31 | 1 | Class C: subnet Class B: subnet Class A: subnet |
Excellent for point-to-point links, see RFC3021 Formerly known as "Useless net" |
4 | 255.255.255.252 | 0.0.0.3 | /30 | 2 | Class C: subnet Class B: subnet Class A: subnet |
"Glue network" (common point to point links) |
8 | 255.255.255.248 | 0.0.0.7 | /29 | 3 | Class C: subnet Class B: subnet Class A: subnet |
Small LANs, NAT pool public address |
16 | 255.255.255.240 | 0.0.0.15 | /28 | 4 | Class C: subnet Class B: subnet Class A: subnet |
Small LANs NAT pool public address |
32 | 255.255.255.224 | 0.0.0.31 | /27 | 5 | Class C: subnet Class B: subnet Class A: subnet |
LANs Small Businesses NAT pool public address |
64 | 255.255.255.192 | 0.0.0.63 | /26 | 6 | Class C: subnet Class B: subnet Class A: subnet |
LANs Small/Medium Businesses NAT pool public address Security DMZs |
128 | 255.255.255.128 | 0.0.0.127 | /25 | 7 | Class C: subnet (half of one) Class B: subnet Class A: subnet |
Large LANs Small/Medium Businesses Large NAT pool public address Security DMZs |
256 | 255.255.255.0 | 0.0.0.255 | /24 | 8 | Class C: 1 Class B: subnet Class A: subnet |
Large LANs Medium Businesses Large NAT pool public address Pre-InterNIC allocation for sites with ~100 hosts or less |
512 | 255.255.254.0 | 0.0.1.255 | /23 | 9 | Class C: 2 Class B: subnet Class A: subnet |
Small enterprise LANs Medium/Large Businesses |
1024 | 255.255.252.0 | 0.0.3.255 | /22 | 10 | Class C: 4 Class B: subnet Class A: subnet |
Enterprise LANs Medium/Large Businesses ISP PoP sites |
2048 | 255.255.248.0 | 0.0.7.255 | /21 | 11 | Class C: 8 Class B: subnet Class A: subnet |
Medium/Large Businesses ISP PoP sites Small ISPs Current Regional Internet Registry (RIR) IP allocation boundary |
4096 | 255.255.240.0 | 0.0.15.255 | /20 | 12 | Class C: 16 Class B: subnet Class A: subnet |
Large Businesses ISP PoP sites Small ISPs |
8192 | 255.255.224.0 | 0.0.31.255 | /19 | 13 | Class C: 32 Class B: subnet Class A: subnet |
Large Businesses NSP PoP sites ISPs |
16384 | 255.255.192.0 | 0.0.63.255 | /18 | 14 | Class C: 64 Class B: subnet Class A: subnet |
Large Businesses NSP PoP sites Small NSPs ISPs |
32768 | 255.255.127.0 | 0.0.127.255 | /17 | 15 | Class C: 128 Class B: subnet (half of one) Class A: subnet |
Large Businesses NSP regions Small NSPs Larger ISPs |
65536 | 255.255.0.0 | 0.0.255.255 | /16 | 16 | Class C: 256 Class B: 1 Class A: subnet |
Large Businesses NSP regions Small NSPs Larger ISPs Pre-InterNIC IP allocation for sites larger than ~100 hosts |
From here, you could zip over to my main gweep pages.
Cheers,
joe
Last updated Mon May 8 13:00:37 EDT 2006