Review For Final

Spring 2011

Network Architecture Models

3OSI Reference Model

7 layers

Each layer communicate with its peer layer, and with layer above and below it.

Different protocols at each layer

Upper layer deal with application issues, and are implemented in software

Lower layers handle data transport issues, and are implemented in software and hardware

App

lica

tion

Dat

a tr

ansp

ort

4Summary Questions

1. (a) Name the 7 layers of the OSI model, (b) Name OSI layers that deal with application issues ? (b) Name the OSI layer that deal with data transport issue ?

Answer: See previous slide

2) Application programs used at the application layer of the OSI Model include word processing. T / F

3) Upper layers of the OSI Model are usually implemented in hardware devices. T / F

Internetworking

6Summary Questions

1. The local area network shown in Figure 8-18 (p. 261) has two hubs (X and Y) interconnecting the workstations and servers. What workstations and servers will receive a copy of a packet if the following workstations/servers transmit a message:

Workstation 1 sends a message to workstation 3:

Workstation 2 sends a message to Server 1:

Server 1 sends a message to workstation 3:

2. Replace hub Y with a switch. Now, what workstations and servers will receive a copy of a packet if the following workstations/servers transmit a message:

Workstation 1 sends a message to workstation 3:

Workstation 2 sends a message to Server 1:

Server 1 sends a message to workstation 3:

7Summary Questions

3. Do Internetworking Exercise 2

The exercise is available in the Notes section. Check the link next to the Internetworking II ppt notes (posted on Thu 2/4)

Fundamentals of Data &Signals

9Summary Questions

1. Distinguish between digital and analog signals

Answer: (see class notes for details if needed): Analog: Continuous waveform, State of signal (voltage

level, etc.) varies continuously among an infinite number of states

Digital: Discrete (non-continuous) waveform, State of signal remain constant during each clock-cycle, Few possible states

2. Distinguish between digital and binary transmission

Answer: Digital: Few possible states (2-64 states)

Binary: 2 states

10Attenuation

A signal starts at a transmitter with 10 watts of power and arrived at a receiver with 5 watts of power. Calculate the loss of power in dB.

1. dB = 10 log10 (P2 / P1)

2. dB = 10 log10 (5/10)

3. dB = 10 log10 (0.5)

4. dB = 10 (-0.3)

5. dB = -3

Q: What is the decibel loss of a signal that starts at 50 watts and experiences a 10-watt loss over a given section of cable ?

Q: What is the decibel loss of a signal that loses half its power during the course of transmission ?

11Frequency, Bandwidth(See Fundamentals of Data & Signals – Continued ppt notes for details if needed)

Q: What is the frequency (in Hz) of a signal that repeats 80000 times within one minute ?

Answer: 80000 / 60 = 1333.33 Hz

Q: What is the bandwidth of a signal composed of frequencies from 50 Hz to 500 Hz ?

Answer: |500 Hz – 50 Hz | = 450 Hz

Conducted and WirelessMedia

13Summary Questions

What does cross talk interference mean ? How can crosstalk interference be reduced ?

If transmission wires are placed side by side, electromagnetic radiation is emitted by one wire and picked up by the other. Twisting pair of wire reduce crosstalk interference.

What categories of twisted pair are usually used in LANs ?

Categories 3, 4, 5, 6, 7, 8

What are the advantages and disadvantages of STP compared to UTP ?

Advantage: better level of isolation from interference. Disadvantage: cost

What is the main difference between Baseband coaxial cable and Broadband coaxial cable ?

Baseband transmit signals in a single channel, whereas broadband uses multiple channels.

What are the advantages of Optical fiber compared twisted pair and coaxial cable ?

Speed, No significant noise, No interference, Long distance

Can you transmit video signal over twisted pair wire ? Explain

Yes, you can (e.g. Digital Video Service or DVS). Noise used to be a limiting factor.

Local Area NetworksLAN Management

School of BusinessEastern Illinois University

15

Summary Questions

What are the four factors to take into account in deciding how many servers to use to implement a LAN’s services?

Answer: Optimization, reliability, security, cost To what two devices does a print server usually connect?

Answer: To a printer via a parallel or USB cable, and to a hub/switch via UTP or other transmission media.

In print service with a print server device, where does a print job go when it leaves the client PC (not counting the hub or switch)?

Answer: The print job first goes to the computer where the print queue manager program is installed, which puts it in a print queue.

Do you have to use special printers for print service?

Answer: You do not need special printers, because all printers come with a parallel and/or USB port that could be used.

16Servers

Single server Versus Multiple specialized Servers– Decision based on Cost, Optimization, Reliability, and Security

Optimization: – File servers need storage capacity and rapid access– Client/Server applications need very fast processors

17Print Service

A Print Server device is basically: NIC + a parallel or USB port + Random Access Memory + Intelligence to receive data and commands from print queue manager program.

Note: Possible to connect printer directly to file server, but people might have to walk far to get their printout.

Network printers have built-in NIC

& print server hardware

18Summary Questions

Answers are in Week8SummaryQuest.doc (Review section of course Website - Check Exam 2 Review link)

Your organization has 12 employees, each with his or her own stand-alone PC running Windows 98. Each computer has a 10 Mbps NIC that could work with coaxial cable or twisted pair.

a) List all the additional hardware and software you would have to buy to install a 100BaseTX LAN. Be very sure that you list all the things the organization will have to buy. The organization wishes to use electronic mail, word processing, file sharing, and print sharing with four existing printers fed with parallel ports.

b) How many ports should the hub or switch have? Explain.

19Ethernet standards

20Summary Questions

1. Directory Applications has Subdirectories Databases and WordProcessings. The Network administrator assigns user Lee to the group Outer. The administrator assigns Outer the access rights R, S, and T in Directory Applications. (Don’t worry about the meaning of R, S, and T. They are simply types of rights.) The administrator assigns Outer the access rights S, U, and V in Subdirectory Databases. a) What access rights does user Lee have in Directory Applications? Explain. b) What access rights does user Lee have in Directory Databases? Explain. c) What access rights does user Lee have in Directory WordProcessings? Explain.

Applications

WordProcessingsDatabases

21Automatic Inheritance of Access Rights

Assigning rights to users or group in a directory Rights automatically inherited in lower directories Simplifies rights assignment

Application

Word Processing Database

Oracle QuickDB

Assigned BrowseAnd Read Rights

Inherits BrowseAnd Read Rights

Inherits BrowseAnd Read Rights

22Automatic Inheritance of Access Rights

Blocking of Inheritance– If rights explicitly assigned in subdirectory, inheritance is blocked– Only assigned rights are effective

Application

Word Processing Database

(Browse and Execute Only)QuickDB

Assigned BrowseAnd Read Rights

Inherit BrowseAnd Read Rights

Assigned BrowseAnd Execute Rights Oracle

The Internet

School of BusinessEastern Illinois University

24The Internet

NetworkBrowser

Packet

Router

Packet

Route

WebserverSoftware

User PC (Host)Operating SystemTCP/IPIP Address =128.150.50.9

Webserver (Host)Operating System

TCP/IPIP Address=139.67.14.54

Host name =eiu.edu

25IP address

32-bits and Dotted Decimal Notation

– IP addresses are really strings of 32 bits (1s and 0s) 10000000101010100001000100001101

– To convert this to dotted decimal notation, first, divide them into four bytes (also called octets)

10000000 10101010 00010001 00001101

(Both octets and bytes are collections of eight bits)

– Convert each binary (Base 2) octet into decimal (Base 10)

26

IP address

Place Value(2N) Bit Decimal

128 1 128

64 0 0

32 1 32

16 0 0

8 0 0

4 0 0

2 1 2

1 1 1

163

Position(N)

7

6

5

4

3

2

1

0

Binary10100011

=Decimal

163

Note: Starts with 0

27IP Address structure

Each IP Address has two main parts:

Each Organization is given the Network Part by an IP address Registrar (e.g: www.arin.net)

For Eastern Illinois University, this is 139.67– All IP Addresses for Eastern’s computers begin with

that Network Part

IP Address139.67

Network Part Local Part

American Registry for Internet Numbers Check EIU’s IP registration info

28IP Address structure

Network Parts is 8 to 24 bits long– For Eastern, it is 16 bits long– 16 bits is only an example

Local Part– Total address is 32 bits, so if the network part is 8 bits, the

local part is 24 bits

Common way to refer to IP address structure:– x.x.x.x/# (where # is the number of bits in the network part)– e.g. 139.67.0.0/16

Network Part(8 to 24 bits)

29Network classes The value of the bits in the first octet of an IP

address determines the Network class

Class AClass BClass C

0.x.x.x to 127.x.x.x128.0.x.x to 191.255.x.x192.0.0.x to 223.255.255.x

Class Address range0xxx10xx110x

Leftmost bits

8 bits16 bits24 bits

Network Part Length

Place Value(2N) Bit Decimal

128 64 32 16 8 4 2 1

Position(N)

76543210

1) For each of the following IP addresses, give the class and the network bits.

10101010111110000101010100000001

01010100111110000101010100000001

2) What class does Eastern’s network belong to? (Net. Part =139.67)

30Host name

Host Names– Like nicknames

Not official addresses Each host must have an IP address But only some hosts have host names If you give it a host name, your browser must look

up IP address of host

eiu.edu

31Domain Name System (DNS)

DNS provides server’s IP addresses given their host names

32Autoconfiguration

User PCs do not need permanent IP addresses– They only need to be found within a use session– They usually are given temporary IP addresses

to use on the Internet for a couple of days– The duration of temporary address is usually a

few days. When the lease expired another temporary address is a given.

Autoconfiguration service provides clients computers with temporary IP addresses to use for communication with other

computers on the Internet.

33Autoconfiguration

Request-Response Cycle– User software requests IP address for the user PC in

Autoconfiguration Request message

– Autoconfiguration Response message contains temporary IP address to use in current session

34Autoconfiguration

Most popular autoconfiguration protocol is DHCP– Dynamic Host Configuration Protocol– Built into Windows after Win 3.1– Supplies host with temporary IP address

DHCP can give more information too– Usually gives IP address of a default gateway

(Microsoft terminology for router)– Can give IP address of a local DNS host– Can give other information

35Summary Questions

1. a) Distinguish between IP address and host name. b) Which is the official address of a host? c) Does a server host need an IP address? d) Does your home PC need an IP address when you are on the Internet? e) Does a server host need a host name? f) Does your home PC need a host name when you are on the Internet?

2. Using the conversion system in slide #26, convert the following IP address to dotted decimal notation: 10101010 11110000 11001100 01010101. (spaces are included to facilitate reading.)

3. Using the conversion system in slide #26, convert the following dotted decimal IP address into a 32-bit IP address: 192.128.2.4

36

Summary Questions

4. What are the two parts in IP addresses?

5. a) Who assigns the Network part? b) The Local part?

6. a) When do we need DNS? b) What information do you send in a DNS request message? c) What information do you receive in a DNS response message?

7. a) What is autoconfiguration? b) What information do we get back, at a minimum, in an autoconfiguration response message? c) What other information may we get back?