INTERNET ADDRESSING

1
INTERNET ADDRESSING
2
Internet Protocol Version 4 (IPv4)

• In 1973, Vint Cerf and Bob Kahn (then working for
  DARPA) decided that a 32-bit address was more
  than adequate for the foreseeable future
  
• Provides for 232 addresses
• 4,294,967,296 IPv4 addresses

3
Robert Shaws Workstations 32-bit Address
10011100001101011000001000100000
IP Addresses often expressed in dotted-decimal
notation made up of four 8-bit (byte) values
10011100.00110101.10000010.00100000
156
.106
.130
.32
4
Internet Scaling Problems

• Ability of Internet backbone routers to route
  traffic between increasing number of networks on
  the public Internet
  
• Potential exhaustion of allocated IPv4 addresses

5
Scaling Ability of Internet backbone Routers to
route traffic

• Router is device/software that determines the
  next network point to which a packet should be
  forwarded toward its final destination
  
• Operates at network layer (layer 3) of the OSI
  Reference Model

6
Scaling Routing Problems

• Routers create/maintain table of the available
  routes and conditions and uses this along with
  distance and cost algorithms to determine the
  best path for a given packet.
• Exponential growth of Internet is reflected in
  growth of global routing tables and computing
  power needed to calculate routing table/topology
  changes

7
Exhaustion of IPv4 address space

• Portions of the IP address space have not been
  efficiently allocated
  
• Original allocation model of classful
  addressing was not efficient

8
Classful IP Addressing

• When IP standardized in Sept 81, each system
  attached to IP network defined to have unique
  32-bit address
  
• Some systems, like Routers, must have interfaces
  to more than one network so must be assigned a
  unique IP address for each network interface
  
• Original idea was that different class size
  would reflect number of hosts in organizational
  networks

9
Classful IP Addressing

• Two-level address structure created

Network-Number
Host-Number
or
Network-Prefix
Host-Number

• First part identifies network on which the host
  resides

• Second part identifies host on the given network

10
Primary ClassfulAddress Classes
11
Class A Networks (/8 Prefixes)

• 8 bit network prefix with highest order set to 0
  followed by 7 bit network number and 24 bit host
  number
  
• Maximum of 126 (27 -2) /8 networks can be defined

12
Class A Networks (/8 Prefixes)

• Each /8 supports maximum of 16,777,214 (224-2)
  hosts per network
  
• /8 contains 231 (2,147,483,648) addresses which
  is 50 of the total IPv4 address space

13
Class B Networks (/16 Prefixes)

• 16 bit network prefix with highest order bits set
  to 1-0 followed by 14 bit network number and 16
  bit host number
  
• Maximum of 16,384 (214) /16 networks can be
  defined

14
Class B Networks (/16 Prefixes)

• Each /16 supports maximum of 65,534 (216-2)
  hosts per network
  
• /16 contains 230 (1,073,741,824) addresses which
  is 25 of the total IPv4 address space

15
Class C Networks (/24 Prefixes)

• 24 bit network prefix with 3 highest bits set to
  1-1-0 with 21 bit network number and 8 bit host
  number
  
• Maximum of 2,097,152 (221) /24 networks can be
  defined

16
Class C Networks (/24 Prefixes)

• Each /24 supports maximum of 254 (28-2) hosts
  per network
  
• /24 contains 229 (536,870,912) addresses which
  is 12.5 of the total IPv4 address space

17
Recognizing Classful Addresses

• A (/8 prefixes)
• B (/16 prefixes)
• C (/24 prefixes)

• 1.xxx.xxx.xxx - 126.xxx.xxx.xxx
• 128.0.xxx.xxx - 191.255.xxx.xxx
• 192.0.0.xxx - 223.255.255.xxx

Where xxx host number
18
Limitations to Classful Addressing

• During early days of Internet, appearance of
  unlimited IP address space allowed allocations
  be made on request and not need
  
• Classful A, B, and C octet did not permit
  efficient allocation of finite address space (our
  shoe sizes are small, medium and large)
  

19
Limitations to Classful Addressing

• Lets take a medium-size organization
• /24 supports 254 hosts
• /16 supports 65,534 hosts
• Which would you choose?
• In past, instead of assigning several /24
  addresss, were given a /16 which has led to
  premature depletion of /16 space
  
• but /24s increase size of routing tables..

20
Subnetting

• In 1985, RFC 950 defined procedure to support
  subnetting or division of a Class A, B, or Cs
  into smaller pieces
  
• To overcome
• Internet routing table growth
• need to request additional network numbers when
  installing new networks

21
Subnetting - More Hierarchy to Classful Addresses
22
Subnetting

• Example site with several logical networks can
  be assigned a single /16 (Class B) network
  address
  
• Traffic forwarded by router to interior subnets
  based on 3rd octet value
  
• Subnet structure is never visible outside of
  private network

23
Subnetting
130.5.32.0 130.5.64.0 130.5.96.0 130.5.128.0 1
30.5.160.0
130.5.192.0 130.5.224.0
Private Network
130.5.0.0
INTERNET
router
24
Subnetting Advantages

• Size of global Internet routing tables does not
  grow
  
• Local administrator can deploy additional subnets
  without obtaining a new network number
  
• Route flapping (rapid changing of routes) within
  private network does not affect Internet routing
  tables

25
Extended Network Prefix

• Internet routers use only network prefix of
  destination address to route traffic to subnetted
  environment
  
• Routers within subnetted environment use
  extended-network-prefix to route between subnets

26
Extended Network Prefix

• Extended-network-prefix identified by subnet
  mask
  
• For example, if you have /16 address of 130.5.0.0
  and you want to use the 3rd octet to represent
  subnet-number, the subnet mask is 255.255.255.0

27
Extended Network Prefix

• Subnet masks bits set to
• 1 if examining system should treat corresponding
  bit as part of extended network prefix
  
• 0 if examining system should treat corresponding
  bit as part of the host number

IP Address 130.5.5.25
10000010.00000101.00000101.00011001
Subnet Mask 255.255.255.0 11111111.11111111.1111
1111.00000000
28
Extended Network Prefix

• Instead of dotted-decimal notation used in
  subnet mask (255.255.255.0), standards now refer
  to extended-network-prefix length
  
• So network address 130.5.5.25 with subnet mask of
  255.255.255.0 can also be expressed as
  130.5.5.25/24

29
Variable Length Subnet Masks

• In 1987, RFC 1009 specified how a subnetted
  network could use more than one subnet mask
  
• When a network is assigned more than one subnet
  mask, it is consider a network with variable
  length subnet masks since extended-network-prefix
  es have different lengths

30
Variable Length Subnet Masks

• VLSM supports more efficient use of an
  organizations assigned IP address space
  
• With only a single subnet mask across a network
  prefix, an organization was locked into a fixed
  number of fixed size subnets

31
Variable Length Subnet Masks

• Example fixed /22 extended network prefix means
  hosts fixed at 210-2 (1,022 hosts)

host-number of bits
Subnet-number of bits
Network Prefix
130.5.0.0/22 10000010.00000101.00000000.00000000
Extended-Network Prefix
32
Variable Length Subnet Masks

• VLSM permits slicing and dicing subnets into
  different sizes and therefore numbers of hosts in
  subnets
  
• VLSM also supports recursive division of an
  organizations address space so it can be
  reassembled and aggregated to hide the complexity
  of routing information at each level

33
Route Aggregation
34
Route Aggregation
35
Classless Inter-Domain Routing (CIDR)

• CIDR documented in Sept 1993 in RFC 1517, 1518,
  1519, and 1520
  
• Eliminates concept of Class A, B, C, classful
  network addresses
  
• Supports route aggregation where single routing
  table entry can represent thousands of classful
  routes

36
Classless Inter-Domain Routing (CIDR)

• Uses generalized concept of network prefix and
  routers use this prefix rather than first 3 bits
  of an IP address to determine dividing point
  between network number and host number
• Supports deployment of arbitrarily sized networks
  (shoes come in all sizes)

37
Classless Inter-Domain Routing (CIDR)

• In classless environment, prefixes are viewed as
  bitwise contiguous blocks of IP address space
  
• For example, all prefixes with a /20 prefix
  represent the same amount of host address space
  (212 or 4,096)

38
Classless Inter-Domain Routing (CIDR)

• Example of /20 (212 or 4,096 hosts) assigned to
  traditional classful Class A, B, and C

Traditional A 10.23.64.0/20
00001010.00010111.01000000.00000000
Traditional B 130.5.0.0/20
10000010.00000101.00000000.00000000
Traditional C 200.7.128.0/20 11001000.00000111.1
0000000.00000000
39
Most Commonly Deployed CIDR Address Blocks