Dr. V.T. Raja

uptightexampleΔίκτυα και Επικοινωνίες

24 Οκτ 2013 (πριν από 3 χρόνια και 11 μήνες)

105 εμφανίσεις

TRANSPORT and NETWORK LAYERS
-

Part 1


Dr. V.T. Raja

Oregon State University

Transport and Network Layers
-

TCP/IP
(Part 1)


Chapter Objectives
:


Understand primary functions of transport and
network layers



Understand the “big picture”


Five
-
layer network model

Examples of Transport/Network Layer Protocols


Examples:


Two popular transport/network layer protocols



Another example of protocol in a WAN environment
that is not popular in North America



Protocol for IBM compatible hardware/software and
mainframe environments

Examples of Transport/Network Layer Protocols


Examples:


Two popular transport/network layer protocols


TCP/IP


IPX/SPX


Another example of protocol in a WAN environment that is not
popular in North America


X.25


Protocol for IBM compatible hardware/software and mainframe
environments


SNA


APPN

TCP/IP


Why TCP/IP?


What is the expansion of TCP/IP?


What are the major functions of TCP/IP?

TCP/IP


Why TCP/IP?


Most widely used protocol on Internet


Estimated about 70
-
80% of LAN, MAN, and backbone networks use
TCP/IP


Reliable protocol


TCP/IP performs some error detection/error correction functions,
although this is not a primary function of the network layer protocol


Compatible with many data link protocols


What is the expansion of TCP/IP?


Transmission Control Protocol/Internet Protocol


What are the major functions of TCP/IP?


Packetizing, Addressing and Routing

TCP


Main functions of TCP:
Packetizing


What is Packetizing?


At sender




At final destination







Thought Questions (with regard to the “big picture”):


With what layer(s) does TCP interface in the 5
-
layer model?


If more than two computers are involved in message transmission, then is
TCP involved in all the computers? In how many computers is TCP
involved?



Transmission Control Protocol (TCP)


Main function of TCP:


Packetizing


What is packetizing?


At sender


Break messages into packets


Assign packet numbers


At final destination


Check if all packets have arrived


Reassemble packets


Interfacing with two other layers


Thought Questions (with regard to the “big picture”):


With what layer(s) does TCP interface in the 5
-
layer model?


If more than two computers are involved in message transmission,
then is TCP involved in all the computers? In how many
computers is TCP involved?

TCP
(Continued)


Some details on TCP header
:


TCP header containing 24 bytes (192
-
bit header) of
overhead information is enclosed along with user’s
data:


Examples of overhead information added:


Source and destination address


Packet sequence number


Thought Question (with regard to the “big picture”):


Didn’t the data link layer already have the source and
destination address specified in the DLH? Is TCP
duplicating this function of DL?



Token Ring and Ethernet Frames


Start Frame Destination Source Message End


delimiter control address address variable delimiter

Token Ring format

Access







Frame

control







check sequence

1 byte







4 bytes

Ethernet format

Destination Source Length Message CRC
-
32

address address 2 bytes


variable


4 bytes

6 bytes


6 bytes

Port Identifiers


Source Port IDs refer to software used at application layer of source
computer


Destination Port IDs refer to software to be used at application layer of
final destination computer


With TCP/IP, each application layer software has a unique port
address/identifier.


Every standard application layer software on the Internet has a
predefined (default) port number.


Thought Questions:


Could network managers choose not to use the default port
numbers?


Why would a network manager not want to use a default port
number?


Internet Protocol (IP)


What are IP’s main functions?


Main functions:




Thought Question:


Is IP involved only at sender’s or final destination
computers? Or is it involved in all interim computers
that take part in the message routing as well?




Internet Protocol (IP)


What are IP’s main functions?


Main functions:


Addressing


Routing


Thought Question:


Is IP protocol involved only at sender’s or final
destination computers? Or is it involved in all interim
computers that take part in the message routing as well?


IP is involved not only at source and final destination, but also
involved at each of the intervening computers involved in the
message transmission.

IP
(Continued)


Different versions of IP


IPv4 and IPv6


IP creates a IP header


IP header contains (among other overhead information)
source and destination address!


Network layer (IP) address


Examples?


Thought Question (with regard to the “big picture”):


In what order are the different headers from different layers combined
with the user data packet?


If DL protocol is Ethernet, and NL protocol is TCP/IP, then what is the
maximum packet size for user data? Assume DLH and DLT sizes sum to
26 bytes, TCP header size is 24 bytes, and IP header size is 24 bytes as
well.


Sample content in each header?

IPv4 Addressing

ICANN


Every computer participating on the Internet
should use approved network layer (IP) addresses.


IP Addresses are approved by ICANN or one of
several private companies authorized by ICANN


Internet Corporation for Assigned Names and Numbers



http://www.icann.org/



FAQ on registration of IP addresses


http://www.internic.net/faqs/


IPv4 Addressing

Dotted Decimal Notation


Dotted Decimal Notation


Four bytes (8 bits = 1 byte) per address


Each byte separated by a dot


Each byte expressed in decimal notation


Example:


Dotted Decimal Notation: 192.16.224.254


Binary Notation?:


What is the minimum decimal value any byte can be assigned?


What is the maximum decimal value any byte can be assigned?


IPv4 Addressing

Class Based Address Structure


IP Addresses were grouped into classes:


Classes A, B, C, D and E


Classes A, B and C
-

used for commercial purposes


The class of the address
-

determined by
examining the first byte of the address and
mapping it to a range of values as follows:


Class A: first byte is in the range 1


127


Class B: first byte is in the range 128
-

191


Class C: first byte is in the range 192
-

223

IPv4 Addressing

Network and Host Addresses


Each IP address has two parts:


Network Address
(assigned/approved by ICANN)


Host Address (assigned by host organization)


For a Class A IP address, the first byte is assigned/approved by ICANN



Class A: Network.Host.Host.Host


For a Class B IP address, the first two bytes are assigned/approved by ICANN



Class B: Network.Network.Host.Host


For a Class C IP address, the first three bytes are assigned/approved by ICANN



Class C: Network.Network.Network.Host



Restrictions for Host address: IP address cannot end in 0 or 255

IPv4 Addressing

Class Participation Exercise # 1


Class Participation Exercise
:

1.
If OSU has an IP address whose first byte is 128, what class
does this IP address belong to?

2.
To which of the four bytes of the IP address can OSU assign
values, assuming the first byte has the value 128?

3.
a) Assume that ICANN has assigned OSU a group of IP
addresses, and that the first two bytes (of each assigned IP
address) are 130.192. In this case, how many IP addresses are
available for OSU?


b) How many IP addresses are available to OSU, if ICANN
assigned OSU a class A IP address?

c) How many IP addresses are available to OSU, if ICANN
assigned OSU a class C IP address?

IPv4 Addressing
(Continued)

Subnets


Subnets


Part of a network that is logically grouped by IP addresses


Example:


Suppose ICANN assigned OSU a set of IP addresses as follows:
128.193.x.x


OSU assigns the host part of the IP address


128.193.75.x; (CS subnet)


128.193.76.x; (BSG subnet)


128.193.76.x; (COB subnets)


Can have one subnet for faculty: 128.193.76.____ (faculty subnet)


Can have one subnet for IS staff: 128.193.76.____ (IS staff subnet)



Every computer in a TCP/IP network is assigned a
subnet mask

by the host organization


Why?

IPv4 Addressing

Subnet Masks


Subnet Mask Applications
:


Helps determine if sender and receiver are on
same subnet.


Why is this important?


Helps determine maximum number of
computers per subnet


Helps determine maximum number of subnets
possible


Subnet mask application

Determining if sender and receiver are on same subnet


How to determine if sender and receiver are
on the same subnet?


Refer to algorithm provided on separate handout


Examples:


Are computer A (IP address: 128.193.78.3) and computer
B (IP address: 128.193.78.250) on the same subnet?


Do you have sufficient information to answer the
above mentioned question?


Class Participation Exercise # 2 (Question # 1)

Subnets
(Continued)

Partial Subnetting


Subnet mask default values


Example: Class Participation Exercise # 2 (Question # 2)


Partial Subnetting



Subnet masks can also take values other than 0 or 255


Example 1:


Subnet mask: 255.255.255.224


Sender’s IP: 120.192.40.16


Receiver’s IP: 120.192.40.30


Example 2: Class Participation Exercise # 2 (Question # 3)

Subnet Mask Applications


Subnet mask helps:


Determine if sender and receiver are on same subnet


Determine maximum number of computers per subnet


Determine maximum number of subnets possible


Example 1:


Background
: Assume OSU is assigned a Class C address, and assume
that the network manager at OSU has selected the subnet mask value as:
255.255.255.248


Question: Find (a) maximum # of computers per subnet and




(b) maximum # of subnets possible

Number of Subnets and Number of Computers Per Subnet

Class C Addresses

Subnet Mask

# of Computers
per Subnet

# of Subnets

255.255.255.224

(3
-
bit subnetting)

255.255.255.240

(4
-
bit subnetting)

255.255.255.248

(5
-
bit subnetting)

255.255.255.255

(CP # 2
-

Q4)

Class Participation Exercise # 2


Question # 5


Background Info
: A firm has been assigned a class B
address by
ICANN
. Based on an estimation of future
network growth in the firm, the administration plans to
have 30 different subnets, and about 2040 computers in
each subnet.


Question
:

What subnet mask value should the network
manager use in order to satisfy the planning needs (of the
administration) as specified above?