The Internet is a hierarchical GLOBAL communication network of interconnected computer networks. At the top of the hierarchy is the backbone network that goes around the globe which is like a huge electronic data pipe (edp) capable of speed, of up to 600 Mbps. Smaller edps branch off from the backbone network and so on down to the home user whose modem connection may have a typical speed of 28.8 kbps or less. The size of the edps is measured in terms of bandwidth - the higher the bandwidth the more the data that can be transported through it.

Leasing of edps is similar to whole sale and retail businesses, the bigger companies acquiring high bandwidth which they in turn divide into smaller bandwidths and lease to smaller companies. Many of the small companies may have dedicated leased lines that are always connected to the Internet and in this way they will have their networks directly and permanently attached to the Internet. Some of the companies called the Internet Service Providers (ISPs), are in the business of providing Internet connectivity to other small institutions and individual users. The end user connects to an ISP with a dial-up modem on making a phone call to the ISP.

The amount of data carried on the Internet is like the amount of water that flows in the city - water distribution system. The bigger the pipe the more the water it carries. The huge pipe brings water to the city boundary and then the pipe sizes are progressively reduced down to the smaller household. The Internet information is divided into chunks and placed into packets, which are then addressed to a destination computer and sent over the Internet. The transportation of the packets is analogous to the freeway system in which the car is an independent delivery system just like the information packet. At the freeway intersection the driver decides which turn to take so as to quickly reach the destination. In Internet, a device called the ROUTER take those decisions at the intersection.

Just like traffic rules, which regulate freeway users, there are also rules that define packet communications and handling. One SET of such rules is the TCP/IP (Transmission Control Protocol/Internet Protocol) which is the most supported protocols it specifically defines:

• how much data goes into a packet;
• how to address the packet;
• how to transmit it over the network;
• how to route around failed links;
• how to detect errors and missing packets and get retransmission.

The Internet provide "products" (resources) to customers (users). What are the products, how are these products transported and how are they made available (accessed) to customers, these will be covered in the next section.

2.3.1     Internet Resources

There are several products from cyberland, which are collectively known as resources, that customers world wide would like to have. These products are information-based, and the Internet is a communication system which facilitates the transpiration and sharing of the products by members of the cybertribe. The cyberland products include:

• Electronic mail
• File Transfers
• The World Wide Web (WWW) or in short "the Web"

The resources have identifiers, normally the name of the protocols that are used in their handling:

• Electronic Maile-mail
• File TransferFTP (File Transfer Protocol)
• WWWHTTP (Hyper Text Transfer Protocol)

To transport any product at times, it would require unique identification marks. In the Internet communication system the identification marks include:

Just like an individual wishing, to make a distant communication a unique address must be used. All communicating devices (PCs, Routers, Printers, etc.) were directly connected to the Internet and MUST also have a unique address. The address is a 32-bit binary number with a special format like:

Human find these numbers difficult to remember. So unique names are used instead, and the Internet system map the names into corresponding numbers. The number/name is called the IP Address. The IP address MUST be unique, no two devices in the whole global Internet system can have the same IP address. It must be emphasised that information communication within Internet is ONLY based on the IP numbers and NOT on the names, we, human use!

The IP addresses (numbers) are divided into classes - Class A through Class E. Classes A to C are commonly used. Class C, whose first octet runs from 192 to 223 can provide a total of 2,097,152 x 254 unique addresses. This might sound a lot of addresses but the hyper Internet growth is depleting the addresses so fast that new addressing schemes are already being researched on.

These are the names used instead of the IP numbers. The naming system is hierarchical in structure. The Domain Name System (DNS) Server does the mapping of names to numbers. When a domain name is specified in an Internet application, a query is sent to the DNS Server that converts the name into an IP number that is then used in all subsequent communications

Suppose it is required to access an FTP resource, to identify it you need to provide:

• resource type:    FTP
• location where the resource is stored:     Host name
• path to the file:    directory/subdirectory/....../filename.ext

A URL is a standard way of identifying a type of Internet resources and its location. It is basically a string of characters consisting of:

(i) Resource type (identified by protocol):

• http:// for WWW
• ftp://for FTP
• Domain Name (IP): server name where resource is stored.
• Path Name: Specification of the file holding the resource, e.g.

where:

http://                         =    resource
www.bigweb.com        =     domain name
/Travel                          =    pathname of the directory, where to find the source file
MyCruise.htm             =     filename of the source file
htm                                  =    file extension to identify file type

Having manufactured the products and properly labelled them, Internet would transport them to outlet points where they would be temporarily stored before they are delivered to customers. In cyberland, the storage places are called SERVERS. This subsection also explains the management arrangement of the servers within the Internet system.

There can be many types of servers as there are different type of resources:

• e-mail Server
• FTP Server
• Web Server

But there may also be special function Servers, such as:

• Domain Name System (DNS) Server
• Windows Internet Name Service (WINS) Server
• Proxy Server
• Print Server, etc.

It should be emphasized that for a server system to work, three principal items are needed:

• A PC that holds the resource called HOST or SERVER
• A PC requesting the resource called CLIENT
• Application Software: Server version on the Host and client version on the client PC.

Note: More than one server can co-exist on a single physical PC.

Recall the term "server", it refers to both the physical location where a resource is stored and also the software that handles the resource. Both of them have the primary aim of making the resource, in question, available to the users. In doing so, primary consideration must be given to the type and speed of the telephone line that access the resource. There are several ways of implementing a server connection to the Internet, web server is an example:

If you can not afford to own a server (PC) and/or the cost of 24-hour dedicated phone line an ISP can be requested:

• For disk space to host your Server

• To physically locate your server (PC) at the ISP site and connect it to ISP LAN; hence your server will be directly connected to the Internet. The server can then be managed by connecting to the ISP and logging into it as the administrator.

In this type of connection, the server PC will be located at own site and a 24-hour connection to the ISP must be maintained. The connection between own site and ISP could be:

• dial-up voice line of at least 19.2 kbps or better
• ISDN 128 kbps or higher
• other digital services, e.g. T1 (at 1.544 Mbps)

Now you need to buy a large chunk of bandwidth from PTT to enable direct connection to a level on the Internet hierarchy higher than the local ISP. Some of the excess bandwidth can be sold to others who can connect their servers via your site.

Note: A server software could include several services, for example a PC on which WinNT Server has been installed can be configured to include:

• WWW Services
• Gopher Services
• FTP Services
• Inter-Service Manager, etc.

This subsection covers on how a customer places an order and takes delivery of the products he needs using the Internet communication system. Presently, there are three basic ways to connect to the Internet to access server information.

The client computer dials into a Service provider and attaches to a computer at the ISP’s site that is connected directly to the Internet. This is terminal mode where the client is not directly connected to the Internet. And more importantly, in this type of connection, TCP/IP is not used. This is called a SHELL ACCOUNT connection.

The client dials up the ISP. Once logged in, the ISP issues the client with an IP address and establishes a low-level protocol (SLIP or PPP) which allows TCP/IP packets to travel across a normal telephone line. In this case, the client becomes a node on the Internet and can run any Internet or web browser.

In this access method, users are attached to the local Internet LAN which in turn is connected to the Internet through a ROUTER. The router manages traffic in both directions. Performance is determined by the telephone/modem connection. Many users operate at 14.4 or 28.8 kbps. User can have other high speed Internet connections through ISDN or other high-speed leased digital lines.