Transcript Document

CIT 742: Network
Administration and
Security
Mohammed A. Saleh
http://ifm.ac.tz/staff/msaleh/CIT742.html
1
Important Information






Lecturer name: Mr. Mohammed A. S.
Find out about my contact details from my personal
website http://ifm.ac.tz/staff/msaleh
For any questions regarding the course, you can write
me an email at any time. Will try to be prompt for
response.
If you need to see me in my office please book an
appointment
by
writing
me
an
email
[email protected]
All notes will be uploaded on the CIT 742 web page,
http://ifm.ac.tz/staff/msaleh/CIT742.html
A hardcopy will be submitted to the class
representative
2
Basic Rules






Try to attend ALL lectures
Sign against your name on the register
Try not to be late, if you have to be late, come in
quietly.
If you miss a class find out what was covered
Material covered in lectures is examinable
The course notes are not meant to be exhaustive, read
the recommended books; Read critically, point out any
errors in the lecture notes.
3
Required Readings
Author
Year
Title
Publisher
Hermachandran
L.
2003
Computer Communications
Network
Charulatha
Tanenbaum A. S.
2005
Computer Networks
Prentice-Hall
Galo M. A.
2002
Computer Communications
and Networking
Technologies
Pacific Grove
Rowe S and
Schuch
2005
Computer Networking
Pearson
4
Recommended Readings
Author
Year Title
Publisher
Kurose J. F
2001 Computer Networking a
top down approach
Pearson
Halsall
2005 Computer Networking
and Internet
McGraw-Hill
Madhulika J.
2002 Computer Networks
BPB
5
Method of Assessment

Student will be assessed through CA (coursework)





Two compulsory tests – Dates will be announced soon!
Tutorials and Assignments – Must attend tutorials!
Final examinations
Coursework carries 40%
Final Exams carries 60%
6
Introduction


High computer sales are accounted for by both private
and business markets.
Where a business has more than one computer, they
are in practice always connected together in a local
area network



These networks mat be advanced therefore are more or less
costly
Question: Why should so much money (and time) be
spent in designing local area networks?
Question: What can you do with a freestanding PC?
7
Terminologies

Local Area Network (LAN)



1.
2.
3.
4.
5.
Set of physically interconnected computers and computer
equipment within a limited area
A LAN may be connected to other LANs
Components that make up the LAN
Workstations
Servers
Cabling
Interconnected components (switches and hubs)
Other resources – printers and cd roms
8
Cont …
9
LANs, Ethernet and Internet



Development of LANs were accelerated by the
development of the Internet and Ethernet
Ethernet developed further into newer versions, Fast
Ethernet and the Gigabit Ethernet
Question: So where did the Internet originate from?




Originated in the American Department of Defense
They developed the ARPA project (Advanced Research Project
Agency)
ARPAnet, and was the first-ever Internet
It had interconnected networks that used the TCP/IP protocol,
which is still used in today’s Internet
10
Classification of LANs
Local Area Networks




A network that is restricted to a single geographical
location.
Encompasses a relatively small area such as an office
building or school
The function of the LAN is to interconnect workstation
computers for the purposes of sharing files and
resources.
It is typically high speed and cheaper to set up than a
WAN
11
Cont …
12
Cont …
Major Characteristics of LANs
 Every computer has the potential to communicate with
any other computers of the network
 High degree of interconnection between computers
 Easy physical connection of computers in a network
 Inexpensive medium of data transmission
 High data transmission rate
Advantages
 The reliability of network is high because the failure of
one computer in the network does not effect the
functioning for other computers.
13
Cont …



Addition of new computer to network is easy.
High rate of data transmission is possible.
Peripheral devices like magnetic disk and printer can
be shared by other computers.
Disadvantages
 If the communication line fails, the entire network
system breaks down.
14
Cont …
Use of LAN
 Followings are the major areas where LAN is normally
used:
1. File transfers and Access
2. Word and text processing
3. Electronic message handling
4. Remote database access
5. Personal computing
6. Digital voice transmission and storage
15
Cont …
Wide Area Networks (WANs)





It is used to describe a computer network spanning a
regional, national or global area.
For example, for a large company the head quarters
might be at Dar Es Salaam and regional branches at
Dodoma, Arusha, Mwanza and Morogoro
They are built so that users from one location can
communicate with users from another location
WANs are slower than LANs
Often require additional and costly hardware



Routers
Dedicated leased line
Complicated implementation procedures
16
Example of WAN
Wide Area Network
17
Cont …

1.
Major characteristics of WAN
Communication Facility

2.
Remote Data Entry

3.
Video conferencing
Updating data remotely
Centralised Information:

Centralised data storage
18
Difference between LAN and WAN
1.
2.
3.
Coverage: LAN is restricted to limited geographical
area of few kilometers. But WAN covers great distance
and operate nationwide or even worldwide.
Connectivity: In LAN, the computer terminals and
peripheral devices are connected with wires and
coaxial cables. In WAN there is no physical connection.
Communication is done through telephone lines and
satellite links.
Cost: Cost of data transmission in LAN is less because
the transmission medium is owned by a single
organisation. In case of WAN the cost of data
transmission is very high because the transmission
medium used are hired, either telephone lines or
satellite links.
19
Cont …
4.
5.
Speed: The speed of data transmission is much higher
in LAN than in WAN. The transmission speed in LAN
varies from 0.1 to 100 megabits per second. In case of
WAN the speed ranges from 1800 to 9600 bits per
second (bps).
Transmission Errors: Few data transmission errors
occur in LAN compared to WAN. It is because in LAN
the distance covered is negligible.
20
Cont …
Metropolitan Area Networks (MANs)






A network that usually spans across the city or a large
campus
Interconnects several local area networks
It uses a high-speed cable for the interconnections
between different LANs
It might be owned or operated by a single organization
Its geographical scope falls between a WAN and a LAN
MANs provides internet connectivity for LANs in the
same metropolitan area
21
Cont …
22
Life Cycle of LANs


Why would an organization decide to have a LAN?
Always have a Business Case


Reason to spend money
Without a BC a LAN may be purchased for the wrong
reasons
23
Cont …
24
Cont …
1.

Analysis
Finding out what the problem is (preliminary study) and
defining it in a specification of requirements




what should be evaluated are different network operating
systems, mail systems, and other applications
choice of hardware components should also be evaluated.
The phase is generally aimed at establishing what the
system should do, not how it should do it.
Other requirements should include response times,
functionality, security, training, hardware, and life
expectancy.
25
Cont …
2.



Design
Determine how the requirements of the specification are
to be met
Breaking down complex projects into more manageable
subprojects.
An extremely important aspect of the documentation of
the network


It must include all the drawings showing the location of hardware
and cables
An overview of the software to be installed on all the
machines must also be prepared here.
26
Cont …
3.

Implementation
This phase involves the physical installation of the
local area network



Cables are run, software is installed, and computers and other
hardware are put in place
Other important aspects of this phase are user training and
starting new routines
Changes may be major hence making it a long process
27
Cont …
4.

Integration and System Testing
The system must be tested




This is a major and comprehensive task
Testing must be given high priority to enhance security


To ensure that the network meets the requirements set out in the
specification
To ensure that the network is stable
Level of data authorization
To tight up security penetration is absolutely necessary
28
Cont …
5.

Operation and Maintenance
Faults may emerge


Serious consequences if there is unauthorized access
Always focus on balancing security and functionality
29
OSI Reference Model


OSI stands for Open Systems Interconnection, and is a
model for data communication
It was created by IS0 -the International Standardization
Organization, and defines seven layers
30
Cont …
•Application
•Presentation
•Session
•Transport
•Network
•Data Link
•Physical

•All
•Away
•People
•Pizza
•Seem
•Sausage
•To
•Throw
•Need
•Not
•Data
•Do
•Processing
•Please
Easy way to remember the 7 layers of the OSI model
31
Cheat Sheet
32
Cont …


The different layers of this model have different
communication tasks
The OSI model is a description of how communication
takes place
33
Layers
1.

Physical Layer
It carries the signals between the parties through a
medium




conducting media such as pairs of wires, coaxial cable, optical
fiber, or radiating media such as radio or satellite links
Depending on the medium the signals can be electrical signals,
light signals or radio waves
The task on this layer is therefore to carry bits on the
chosen medium.
It also defines other physical characteristics of the
network i.e. identifies which physical topology is to be
used.
34
Cont …
2.





Data Link Layer
It administers the use of the physical layer, in terms of
channel access.
It gathers the bits into blocks of frames, this is the
protocol data unit at layer 2
It offers flow control and applies error control and
corrections
It is also responsible for hardware addressing of frames
It is subdivided into two main sub-layers:


MAC layer: The MAC address is defined at this layer. The MAC
address is the physical or hardware address burned into each
NIC.
LLC layer The LLC layer is responsible for the error and flowcontrol mechanisms of the data-link layer
35
Cont …
3.

Network Layer
With the two previous layers, these problems have
been solved





Conveying signals between the sender and receiver
Flow control
Freedom of errors
We start interconnecting devices to form a network and
interconnect network to other networks, more
problems arise
Imagine you are connecting to a web server, which is
far away, there is no direct line from your PC to the
server
36
Cont …
37
Cont …


The network layer is responsible for carrying data
through a network
Two things are needed:




Every device connected to the network must have a network
address
The routers that connect the network together must have
tables that link the physical lines with possible addresses.
It uses network addresses and internal tables to decide
the route that the data should take through the
network to reach the address of the receiver this is
known as routing
The data on this level is referred to as packets or
datagrams
38
Cont …


4.

Chosen routes may either be dedicated routes or not.
A dedicated route is known as connection-oriented
network, while a route that is not set up is referred to
as connectionless network.
Transport Layer
Known as end-to-end protocol




establishes a connection between the endpoints
Flow control and error control mechanisms are also
performed in layer 4
Error control for the arriving packets
It is also responsible for sequence control, making sure
that packets are delivered in the correct sequence
39
Cont …
5.



6.


Session Layer
This layer is used to establish, monitor, and terminate
sessions between the end points.
The most important task here is the ability to
synchronize the communicating parties.
It does this by establishing, maintaining, and breaking
sessions.
Presentation Layer
This layer ensures that the data is displayed correctly
on the equipment to which it is sent.
Functions such as character conversion, display control,
encryption, compression, and so on are performed
here.
40
Cont …

Some common data formats handled
presentation layer include the following:



by
the
Graphics files JPEG, TIFF, GIF.
Text and data: translate data into different formats such as
American Standard Code for Information Interchange (ASCII)
and the Extended Binary Coded Decimal Interchange Code
(EBCDIC).
Sound/video MPEGs, QuickTime video, and MIDI files
41
Cont …
7.




Application Layer
The function of the application layer is to take requests
and data from the users and pass them to the lower
layers of the OSI model
Incoming information is passed to the application layer,
which then displays the information to the users
The most common misconception about the application
layer is that it represents applications that are used on
a system such as a Web browser, word processor, or a
spread sheet
It defines the processes that enable applications to use
network services, like web, email, file transfer,
terminal emulation programs
42
TCP/IP Model

When the ARPAnet was invented, it resulted to the
following:



Protocols used on the internet are known as the TCP/IP
suite



The IP protocol became the standard for the network
the network came to be known as the Internet
Includes all
standards.
protocols
that
are
recommended
network
The layering changed to have only four layers, now
referred to as the TCP/IP model.
Simpler than the OSI model
43
TCP/IP Model
44
TCP/IP Protocol Suite
45
PDUs and Encapsulation

Every layer also has a protocol (a set of rules)for the
exchange of data units.
46
Cont …



Note that the rules operate horizontally, between peer
layers, while the data is sent vertically between the
layers until it reaches the transport medium
An application protocol defines a set of possible
protocol data units in order to send data or commands
between a client and a server
The protocol of the application now specifies in detail
what this PDU must look like.
47
Cont …

On transmission from client to server, this A-PDU is
handed down to the layer below -in this case TCP
48
Cont …

Note that the application has put a header on the PDU


The TCP layer now regards the A-PDU as data to be
transported



This header says something about the nature of the content of
the PDU, which the receiving side needs to know for correct
processing.
The transport layer adds its own header to manage the
transport service
The same procedure is repeated with the network
layer, where the T-PDU is encapsulated in a network IP
datagram, or N-PDU.
The IP datagram (the N-PDU) is encapsulated in an
Ethernet frame (LPDU),and the network card sends this
frame out on to the medium.
49
Cont …

At the receiving computer, the reverse process takes
place: every layer recovers its PDU by unpacking the
data field from underlying layer
50
Exercises
1.
2.
3.
List the five phases of the waterfall model.
Relate each of the phases to the installation of a local
area network.
Outline the main advantages –and disadvantages-of
installing a local area network in an organization.
51
Topologies

It referrers to the design of LANs

The way LANS are physically set up.

Common topologies are bus, star and ring
1.
Bus topology
Nodes are connected via a long cable.
All the nodes are connected to the same cable, at
different positions.


52
Cont …


An example of this design is a telephone network.
Problems:

Capacity problem





the fact that several different units are sharing a common medium(the
telephone cable)leads to conflicts between the units
Since only one unit at a time can use the cable, capacity could become a
problem
It is defined as collisions
The access method used is CSMA/CD
Security problems


All the nodes that are connected to a bus will be able to hear all
communication that is being carried on the cable
Access to the cable gives access to all the information being carried on this
part of the network.
53
Cont …



The bus topology is a typical example of a shared
network.
Therefore an Ethernet network is a classical example of
a shared network.
How can a message reach the appropriate receiver in a
network with several stations, when every node is able to
read the message?
54
Cont …
2.


Ring Topology
Commonly used method in local area networks.
Technology used is Token Ring
55
Cont …

Data will be sent from one station to another in a welldefined order




It will need to go through several other "neighbours" to reach
its destination.
Any node that needs to transmit will have a ‘token’, which
circulates on the ring
The ring topology does not look like a ring, but the design is
just logical
A hub will normally take care of the connection, Such a
special hub is called a Multistation Access Unit (MAU)
56
Cont …
57
Cont …
3.


Star Topology
Communicate in full through a central node in the
network.
Modern network connect nodes to a switch



The switch becomes the central unit managing traffic between
the other connected units
A switch only sends data to the port for which it is designated
Modem network cabling is in practice only carried out
with structured cabling

everything is put in a star structure
58
Cont …
59
Access Methods



Different topologies can be used for LANs
Some of these assume that several users will be using
the same medium
Common access methods are:


Carrier Sense
(CSMA/CD)
Token passing
Multiple
Access
with
Collision
Detection
60
Contest CSMA/CD



A distributed principle which is used in Ethernet – bus/
tree topology
It stands for Carrier Sense Multiple Access/Collision
Detection
“Carrier Sense”



Means that a device that wishes to send on the medium senses
whether there are other devices sending a carrier at the same
time
If the medium is in use, the station does not send, but waits
until the medium is free
When the medium becomes free it will start its frame on the
medium
61
62
Cont …

“Collisions Detect”


In the event of a collision the parties stop sending and try again
after a random wait
In this way, many users can use the same medium (Multiple
Access).
63
Token Passing



The principle of token passing can be used on both ring
and bus networks
It uses the principle of a token to transmit frames
across the network
Tokens circulate between the participants if the
network is a ring

If the network is a bus, the tokens "circulate” in a particular
sequence



Think of the baton in a relay race
Only the one who "is holding” the token at a given instant can
use the medium
This ensures that only one recipient uses the medium at a time.
64
Cont …

Principle advantages over CSMA/CD




CSMA/CD decreases in efficiency whenever there is a high load
on the network, whereas token passing is always efficient (the
medium can be exploited fully)
The efficiency of CSMA/CD declines sharply when there are
many collisions and the frames have to be re-sent
Token passing offers the ability to prioritize traffic
It can also guarantee a maximum waiting time before a
station can send
65
LAN Standards

The Institute of Electrical and Electronic Engineers
(IEEE) developed a series of networking standards.



to ensure that networking technologies developed by respective
manufacturers are compatible.
cabling, networking devices, and protocols are all
interchangeable when designed under the banner of a specific
IEEE standard.
Each of these IEEE specifications outlines specific
characteristics for LAN networking including:




Speed
Topology
Cabling and
Access method
66
67
802.2 IEEE Standard




Referred to as the Logical Link Control (LLC)
Manages data flow control and error control for the
other IEEE LAN standards.
Data flow control regulates how much data can be
transmitted in a certain amount of time
Error control refers to the recognition and notification
of damaged data.
68
802.3 IEEE Standard





Defines the characteristics for Ethernet networks.
Ethernet networking is by far the most widely
implemented form of local area networking .
Since the development of the original 802.3 standards,
there have also been several additions that have been
assigned new designators.
These standards are often referred to as the 802.3x
standards
Some of the newer standards include:



802.3u for Fast Ethernet
802.3z for Gigabit Ethernet
802.3ae for 10-Gigabit Ethernet
69
Features of IEEE 802.3

Speed


The original IEEE 802.3 standard specified a network transfer
rate of 10Mbps
Modifications to the standard led to:




Fast Ethernet (802.3u), which can transmit network data up to 100Mbps and
higher.
Gigabit Ethernet (802.3z), which can transmit at speeds up to 1000Mbps.
802.3ae is a very fast 803.3 standard. Known as 10-Gigabit Ethernet, it
offers speeds 10 times that of Gigabit Ethernet.
Topology


The original Ethernet networks used a bus or star topology
IEEE 802.3u and 802.3z use the star topology.
70
Cont …

Media





Refers to the physical cabling used to transmit the signal around
the network
The original 802.3 specifications identified coaxial and twisted
pair cabling to be used
The more modern standards specify twisted pair and fiber-optic
cable
802.3ae currently only supports fiber media.
Access method


The access method refers to the way that the network media is
accessed
Ethernet networks use a system called Carrier Sense Multiple
Access with Collision Detection (CSMA/CD).
71
802.5 IEEE Standard



This standard specifies the characteristics for Token
Ring networks.
Developed by IBM in the mid 80’s replaced by the
Ethernet networks.
It is unlikely that you will encounter a ring network in your
travels and even more unlikely that you will be
implementing a ring network as a new installation.
72
Features of IEEE 802.5

Speed


Topology


Token Ring networks use a logical ring topology and most often
a physical star
Media


The 802.5 Token Ring specifies network speeds of 4 and
16Mbps.
Token Ring networks use unshielded twisted pair cabling or
shielded twisted pair.
Access Method


specifies an access method known as token passing
On a Token Ring network, only one computer at a time can
transmit data
73
Fiber Distributed Data Interface



The American National Standards Institute (ANSI)
developed the Fiber Distributed Data Interface (FDDI)
standard
It was developed to meet the growing need for a reliable
and fast networking system to accommodate distributed
applications.
FDDI uses a ring network design


unlike the traditional 802.5 standard, FDDI uses a dual ring
technology for fault tolerance
Because of the dual ring design, FDDI is not susceptible to a
single cable failure like the regular 802.5 IEEE standard
74
Features of IEEE FDDI

Speed


Topology


FDDI uses a dual ring topology for fault-tolerant reasons.
Media


FDDI transmits data at 100Mbps and higher.
FDDI uses fiber-optic cable that enables data transmissions that
exceed two kilometer
Access Method

Similar to 802.5, FDDI uses a token-passing access method.
75
IEEE 802.11 Standard


Specify the characteristics of wireless LAN Ethernet
networks
There are four common wireless standards:


802.11, 802.11a, 802.11b and 802.11g
Each of these wireless standards
characteristics
identifies
several
Features of 802.11
 Speed


Media


measured in Mbps and vary between network standards.
use radio frequency (RF) as a transmission media
Topology

wireless standards can be implemented in an ad-hoc or
76
infrastructure topology
IEEE 802.11 Standard

Access Methods





Uses Carrier Sense Multiple Access/Collision Avoidance
(CSMA/CA).
Variation on the CSMA/CD access method
CSMA/CA access method uses a "listen before talking" strategy
Any system wanting to transmit data must first verify that the
channel is clear before transmitting, thereby avoiding potential
collisions.
Range


Wireless standards each specify a transmission range
It is influenced by many factors such as obstacles or weather.
77
Cont …



IEEE 802.11offered 1 or 2Mbps transmission speeds
802.11a specified speeds of up to 54Mbps
IEEE 802.11b The 802.11b standard provides for a
maximum transmission speed of 11Mbps
78
Questions