TCPIP Overview I

1
TCP/IP Overview (I)
2
TCP/IP Overview

• Trace back to the U.S. Governments ARPA
  (Advanced Research Projects Agency) as early as
  1969
• Two standards adopted by the DoD
• TCP Transmission Control Protocol
• IP Internet Protocol
• Forms the foundation of the Internet as we know
  today
• Contains many user applications, including
  services to
• Electronic Mail (SMTP)
• Remote login (Telnet)
• File transfer (FTP)
• IAB (Internet Architecture Board) is responsible
  for the TCP/IP standards

3
The TCP/IP Protocol Suite

• Consists of four layers
• Network access layer Provides physical delivery
  of IP packets via frames or cells
• Internet layer Contains the information so that
  data can be routed through an IP network
• Host-to-host layer Services the process layer
  and Internet layer to handle reliability and
  session aspects of the transmissions
• Process layer Application support

4
The TCP/IP Protocol Suite
5
Network Access Layer

• TCP/IP relies on a physical network to deliver
  its packets (can be a LAN, a MAN, or a WAN)
• Characteristics
• Process data down from the TCP/IP stack to build
  and send frames or cells out to the network
• Sends frames or cells over the physical network,
  one bit at a time
• Process data up to the TCP/IP stack on the
  receiver side
• Example Ethernet, ATM, and Frame Relay

6
Network Access Layer
7
Internet Layer

• The Internet layer handles several jobs
• Network addressing (ARP RARP) RARP is used to
  find the IP address when the MAC address is known
• Routing information (OSPF, EGP)
• Data fragmentation If a large datagram is
  transferred from network to network, the router
  may divide the datagram into fragments. Each
  fragment has an identification number.
• Reassembly
• Handling error and request (ICMP) When a
  fragment fails to arrive or is corrupted, ICMP
  generates an error message. Also, it allows one
  to see if there is a physical connection to a
  host (ping)

8
Internet Layer
9
IP Header (Version 4)
10
The Version 4 IP Header

• IP version (4 bits)
• IP header length (4 bits)
• Precedence 3 bits
• 000 routing datagram
• 001 priority
• 111 allow diagnostic and management packets to
  go through the network when it is congested or
  having problems
• Type of service (5 bits) (some routers do not
  implement this)
• 00000 normal service
• 10000 low delay request (Telnet)
• 01000 high throughput request (SNMP)
• 00100 high reliability request
• 00010 low monetary cost

11
The Version 4 IP Header

• Total IP length, including the IP header (2
  bytes)
• Datagram ID number (2 bytes)
• Fragmentation (2 bytes with 3 flag bits)
• 000 fragmentation allowed and no more fragments
  to follow
• 001 fragmentation allowed and more fragments to
  follow
• 010 fragmentation not allowed
• Time to live The number of seconds that a
  datagram can live before it must be delivered or
  discarded
• Protocol The ID of the higher layer protocol
  (TCP6, UDP17)
• IP Header check sum
• Source IP address
• Destination IP address

12
IP Addressing (I)

• Dotted decimal notation format
• 4 decimal number separated by decimal points
• Each decimal is one byte in length
• The decimal numbers are in the range 0-255
• Consists of two parts
• The network portion Internet administered
  (cannot be modified)
• The host portion Locally administered (can be
  modified)
• IANA (Internet Assigned Numbers Authority) and
  ARIN (American Registry for Internet Numbers)
  manage the assigning of IP addresses
• Look up who has which address www.arin.net/whois

13
IP Addressing (II)

• Address assignment characteristics
• Addresses are assigned to one of the three
  classes A, B, and C
• Class D is reserved for multicast address
• Class E is reserved for experiment
• Address 127.0.0.0 is used for IP loopback testing
• Addresses reserved for private addresses
• Class A 10.0.0.0
• Class B 172.16.0.0 thru 172.31.0.0
• Class C 192.168.0.0 thru 192.168.255.0

14
IP Addressing (III)

• Class A
• 7 bits for network number, 24 bits for host
  number
• The first bit of the first octet (byte) is a 0
• The first decimal number must fall in the range
  between 0 and 127
• 128 (27) possible networks
• 224 16,777,216 number of hosts per network

15
IP Addressing (IV)

• Class B
• 14 bits are used for network number and 16 bits
  for host numbers
• The first two bits of the first octet begin with
  binary 10
• The first decimal number must fall in the range
  between 128 to 191
• 16,384 (214) possible networks
• 216 65,536 number of hosts per network

16
IP Addressing (V)

• Class C
• 21 bits are for the network number and 8 bits are
  for the host numbers
• The first three bits of the first octet begin
  with binary 110
• The first decimal number must fall in the range
  between 192 to 223
• 2,097,152 (221) possible networks
• 28 256 number of hosts per network

17
Subnetting (I)

• Subnet A network that is a portion of a larger
  network, connected by router
• Why Subnetting
• Performance problems
• Security issues (sensitive data)
• Connectivity issues (different floors, buildings,
  states, or countries)
• Connecting dissimilar media protocols (Ethernet,
  token ring, FDDI)
• Subnetting is accomplished by applying a subnet
  mask, which identifies which bits are used to
  indicate the different portions of the address
  network, subnet, and host

18
Subnetting (II)
19
Subnet Mask

• The mask indicates how many of the host bits have
  been used for subnetting
• The mask is constructed by placing a 1 in any bit
  that is part of the network or subnetwork portion
  of the address
• The mask needs to be configured on all hosts on
  the network

20
Class C Subnetting Example A (I)

• When no subnets are introduced, one network can
  have 256 host numbers (but two cannot be used)
• Subnet mask for no subnets

255
255
255
0
Network Address
Host
21
Class C Subnetting Example A (II)

• 2 subnets are needed
• Use 2 bits for subnet addressing
• The number of bits (x) needed 2x gt the number
  of desired subnets 2 (two addresses cannot be
  used)

194
194
14
Network Address
Subnet Address
Host
22
Class C Subnetting Example A (III)

• First subnet Address 194.194.14.64
• Second subnet address194.194.14.128
• Subnet Mask 255.255.255.192

23
Class C Subnetting Example A (IV)

• Number of available addresses for hosts in each
  subnet 2h 2, where h is the number of bits
  assigned for host
• In our example, h6 and the number of available
  addresses for hosts 26 -2 62
• Total number host addresses 2 subnets 62124
• Before subnetting the number of available host
  addresses 254 (256-2)
• The number of lost addresses130, due to
  subnetting

24
Class C Subnetting Example A (V)

• IP addresses for hosts in the first subnetwork

25
Class C Subnetting Example A (V)

• IP addresses for hosts in the 2nd subnetwork

26
Class C Subnetting Example B

• 196.196.15.0
• Subnet mask without subnetting
• Number of available IP addresses for host without
  subnetting
• Divide the network into 8 subnetworks
• How many bits are needed for subnet addressing?
• What are the eight network addresses?
• How many IP addresses are available for the first
  subnet? List all host IP addresses.
• How many total IP addresses for this
  organization?
• How many IP addresses are lost due to subnetting?
• What is the subnet mask?

27
Class B Subnetting Example

• 130.128.0.0
• Subnet mask without subnetting
• Number of available IP addresses for host without
  subnetting
• Divide the network into 254 subnetworks
• How many bits are needed for subnet addressing?
• What are the eight network addresses?
• How many IP addresses are available for the first
  subnet? List all host IP addresses.
• How many total IP addresses for this
  organization?
• How many IP addresses are lost due to subnetting?
• What is the subnet mask?

28
Class A Subnetting Example

• 3.0.0.0
• Subnet mask without subnetting
• Number of available IP addresses for host without
  subnetting
• Divide the network into 254 subnetworks
• How many bits are needed for subnet addressing?
• What are the eight network addresses?
• How many IP addresses are available for the first
  subnet? List all host IP addresses.
• How many total IP addresses for this
  organization?
• How many IP addresses are lost due to subnetting?
• What is the subnet mask?

29
Variable Length Subnetting (I)

• Consider the following situation
• A Class C network address 200.200.200.0
• One home LAN with 50 hosts
• 3 branch offices, with 20 hosts each

30
Variable Length Subnetting (II)

• For the home office
• 2 bits are assigned for subnet addressing
• The remaining 6 bits are for hosts at the main
  office, allowing up to 62 hosts
• Assume we choose 10 for the subnet address bits
  (we can use 01, too)
• The hosts at the home office are assigned IP
  addresses from 200.200.200.129 to 200.200.200.178
• The last octet for the first and last IP
  addresses are 1000 0001 and 1011 0010
• Subnet Mask 255.255.255.192

31
Variable Length Subnetting (III)

• For the branch offices
• 3 bits are used for subnet addressing
• 001 for Branch 1
• 010 for Branch 2
• 011 for Branch 3
• 100 Used by the main office
• 101 used by the main office
• 110 used by the main office
• IP addresses for the hosts in Branch 1
  200.200.200.33 200.200.200.62
• IP addresses for the hosts in Branch 2
  200.200.200.65 200.200.200.94
• IP addresses for the hosts in Branch 3
  200.200.200.97 200.200.200.126
• The subnet mask for these three branches
  255.255.255.224
• Requires a routing protocol that transmits the
  subnet mask when exchanging routing tables