Classful Addressing and Subnetting

1
Classful AddressingandSubnetting

• Richard Gnall

2
Class A Network

• First bit of first octet is set equal to 0

Host Address
Host Address
24 bits
24 bits
00000000 - 01111111.X.Y.Z
0 - 127.X.Y.Z
7 free bits
7 free bits
Network Address
Network Address
3
Class B Network

• First two bits of first octet is set equal to 10

Host Address
Host Address
16 bits
16 bits
10000000 - 10111111.X.Y.Z
128 - 191.X.Y.Z
14 free bits
14 free bits
Network Address
Network Address
4
Class C Network

• First three bits of first octet is set equal to
  110

Host Address
Host Address
8 bits
8 bits
11000000 - 11011111.X.Y.Z
192 - 223.X.Y.Z
21 free bits
21 free bits
Network Address
Network Address
5
Class D Network

• First four bits of first octet is set equal to
  1110

Host Address
Host Address
8 bits
8 bits
11100000 - 11101111.X.Y.Z
224 - 239.X.Y.Z
20 free bits
20 free bits
Network Address
Network Address
6
Class E Network

• First four bits of first octet is set equal to
  1111

Host Address
Host Address
8 bits
8 bits
11110000 - 11111111.X.Y.Z
240 - 255.X.Y.Z
20 free bits
20 free bits
Network Address
Network Address
7
Network Classes
8
Reserved Address Ranges

• 10.X.Y.Z Private Intranet
• 172.16-32.Y.Z Private Intranet
• 192.168.Y.Z Private Intranet
• 169.254.Y.Z APIPA
• 224 - 239.X.Y.Z Class D Multicasting
• 240 - 255.X.Y.Z Class E Research

Non-Routable on the Public Internet
9
Restricted Addresses

• Network ID/Number
• First IP address in a subnet range
• All host bits set to 0
• Default Gateway (Router) Address
• Usually, but not necessarily, the second IP
  address in a subnet range
• Specifies where to send packets destined for a
  different subnet
• If the subnet is not connected to a router, this
  IP address does not need to be reserved for a
  gateway
• Broadcast Address
• Last IP address in a subnet range
• All host bits set to 1
• Packet will be read by all hosts
• Loopback Address
• 127.X.Y.Z
• 127.0.0.1 ? implies no 127.X.Y.Z
• Messages don't go out on the network, but instead
  just test the installed IP software
• All pings sent to any 127.X.Y.X address get sent
  to 127.0.0.1

10
Subnet Masks

• Additional piece of information identifying the
  network and host sections of the IP address
• A binary 1 for every bit that belongs to the
  network portion of the address.
• A binary 0 for every bit that belongs to the host
  portion of the address.
• IP Address 10110110.11011011.01011010.00110011
• Subnet Mask 11111111.11111111.00000000.00000000
• Subnet Mask 255.255.0.0
• CIDR Notation /16
• Network ID 10110110.11011011.00000000.00000000

11
Subnet Masks
12
Are Subnet Masks Redundant?

• Why bother, if only class A, B and C networks are
  allowed? Isn't it obvious which part of the IP
  address is the network address (by examing the
  first one to four bits of the IP address)? There
  should just be three choices
• 255.0.0.0, 255.255.0.0, or 255.255.255.0
• A network can be broken up into smaller networks
  (subnets) by borrowing some of the host bits to
  help denote a network address. Instead of using
  all or none of the bits in a single octet to
  define the network, a subset of the bits can be
  used to define the network.
• Example
• To subdivide a Class A network into two subnets,
  use the mask
• 11111111.10000000.00000000.00000000 255.128.0.0 /9
  
• To subdivide a Class A network into four subnets,
  use the mask
• 11111111.11000000.00000000.00000000 255.192.0.0 /1
  0

13
Are Subnet Masks Redundant? (cont.)

• To subdivide a Class A network into two subnets,
  use the mask
• 11111111.10000000.00000000.00000000 255.128.0.0 /9
  
• If the class A network is 50.X.Y.Z, then the two
  subnets would be
• 50.0-------.Y.Z
• 50.1-------.Y.Z
• To subdivide a Class A network into four subnets,
  use the mask
• 11111111.11000000.00000000.00000000 255.192.0.0 /1
  0
• If the class A network is 50.X.Y.Z, then the four
  subnets would be
• 50.00------.Y.Z
• 50.01------.Y.Z
• 50.10------.Y.Z
• 50.11------.Y.Z

14
RFC Restricted Subnet Values

• Prior to 1995, RFC 950 ("Internet Standard
  Subnetting Procedure") disallowed the use of all
  0's or all 1's within a single octet to denote
  subnets. It was felt that some routers would be
  confused by this type of subnetting.
• So, using a single bit to subnet a Class A, B, or
  C network is pointless as the first subnet is all
  0's (a single 0) and the second subnet is all 1's
  (a single 1).
• Using 2 bits to subnet a network, results in two
  useful subnets, 01 and 10. The first and the last
  are disallowed, 00 and 11.
• In general, prior to 1995 and under the rules of
  RFC 950, n bits gives 2n - 2 usable subnets.
  (This is important for CISCO certification
  exams.)
• In 1995, RFC 1812, titled "Requirements for IP
  Version 4 Routers", lifted this restriction.
• In general, after 1995 and under the rules of RFC
  1812 , n bits gives 2n usable subnets.
• Note however, that both Microsoft and CISCO
  TCP/IP certification exams follow the earlier RFC
  950 convention.

15
RFC Restricted Subnet Values

• RFC 950/CCNA Subnet Restrictions
• The first and last subnets in a subnetted network
  can not be used. So if a network has been
  subdivided into two subnets, there are in effect
  no available subnets. If a network has been
  subdivided into four subnets, there are in effect
  only two available subnets.
• In each octet, the bits in the portion of the
  network address which represent the network ID
  can not be all 0's or 1's. This slightly reduces
  the number of available subnets.
• ____.____.____.00-------- X not allowed
• ____.____.____.01-------- allowed
• ____.____.____.10-------- allowed
• ____.____.____.11-------- X not allowed
• Technically speaking, the first and last subnets
  would work fine with modern routers, but for
  Microsoft and Cisco certification exams, the
  first and last subnets are not allowed.