How The Web Works

The Internet is not a single network. It is a vast, earth-span-ning network of networks. As you’re reading this, dozens of
networks and PCs are in the process of hooking themselves
up to this international network, extending it almost infinitely.
You might just be pining, craving, desperately wishing that
somebody would come forward and explain how this whole blessed
system works. And that signifies the first step towards geekhood—
and if you read on, this chapter could expedite the process.

The Structure of the Internet:

Webs In General:

Spiders made webs before we did, you know. At first, a spider

makes a bridge line—the first thread that connects two solid

points, like two branches or walls or trees. He then reinforces this

thread with a few more threads. This primary line is what holds

the entire web together. He then proceeds to make a Y-shaped radi-al—a framework—and then the spokes that constitute a web.

If you thought it was just some bright copywriter’s idea to call

the Internet a “web,” think again. There are some remarkable sim-ilarities between how spiders make their webs and how the

Internet takes shape. The primary resemblance is what is known

as the Internet “backbone."

The Central—But Not Really Nervous—System

Akin to the spider’s bridge line is what is called a “backbone,” a

super-fast network that goes all across the world, usually from one

metropolitan area to another. Mega- ISPs run these lines that can

transport data at approximately 45 MBps (on T3 lines) and are linked

at specified interconnection points (called “national access points”

and sometimes known as “regional nerve centres”). All regional net-works are connected to each other by high-speed backbones, which

are basically connections

that can transport data at

very high speeds. When

data is sent from one

regional network to

another, it is first sent to

the above-mentioned NAP

which in turn routes it to

a backbone. The ISP that

connects your computer

to the Internet is a local

one that uses routers to

send and receive data via

this backbone.

About A Router

A router (the “rou” in “router” rhymes with “cow”) is an electron-ic device which comes in a little box, and which connects one net-work with another (for example, a Local Area Network with a

Wide Area Network). Routers do most of the task of directing traf-fic on the Internet by sending the request from your network to

another while simultaneously preventing other unauthorised

users to access your network (this latter task is done with the help

of a firewall, a software program that intercepts and filters

unwanted packets of data that are either trying to enter your net-work or access the resources of your network). A router examines

each packet of data that travels across the Internet, figures out

where it is headed, and “routes” it towards its destination in the

most efficient way possible. Usually, the packet of data goes to

another router and another, and another—until it reaches the

destination. Each router that a packet of data passes through to

reach its destination is considered a “hop.” A router must depend,

however, on gateways. 

 Gateways And Bridges

Just when you—finally—figured out that “windows” aren’t some-thing you open to let fresh air in, these computer guys tell you to

start using “gateways,” too. And you got it right this time—a “gate-way” is not something you open to let people in.

A gateway is a network point—a node—that acts as an entrance

to another network. It is basically a device that allows the

exchange of data by translating between the protocols of two dif-ferent networks or computers on the connected networks. In the

case of your residential connection, the gateway is your ISP.

Gateways are different from Bridges in the sense that a bridge only

links LANs together, sending data to another LAN while keeping

local data within its own network.

Along the way, the data we transact—apart from passing

through several routers, gateways, and backbones—is also depend-ent on an indispensable device called a Server. Servers are of dif-ferent types, and they perform several important functions.

Servers

A server is basically a program that runs on a machine and pro-vides a specific service to all machines that are connected to it. The

server has a specific set of programs that allows the connected

computers (called clients) to make different kinds of requests to it. 

Think of a server as a waiter in a restaurant standing by to give

you all he can accommodate on a tray. If, however, he doesn’t have

the item you wish to consume, he re-routes your order to another

waiter (with different goodies) who happens to be waiting at

another table. This other waiter passes on the object of desire via

other waiters so it finally ends up on your plate.

A Web server is a machine that hosts Web sites and allows

Internet users (clients) to access these sites. A Web server is basically

the kind of server that holds static companies and just waits around

for requests to come around from other machines on the Internet.

Once it receives a request, it allows access to the data requested. 

Web servers are the principal part of the Web, and every URL you

request usually ends up at a Web server—unless a proxy server decid-ed that it can help speed up the process by sending you the data itself.

❍ A file server is a machine that keeps files and allows clients to

upload and download them from it.

❍ A proxy server sits between a client application, such as a Web

browser, and a real server. It intercepts all requests to the real

server, sees if it can fulfil the request from any of the recent

requests, and if it can’t, forwards the request to the server that

actually hosts the Web site. 

❍ A groupware server is software designed to enable users to col-laborate, regardless of location, via the Internet or a corporate

intranet and to work together in a virtual atmosphere.

❍ Mail servers move and store mail all across the Internet, and you

have been “served” by one when you’ve checked your mail.

❍ A chat server allows you to interact with other users on the net-work in real-time.

❍ Integral to the process of sending data across the Internet is the

Domain Name Server (“the DNS”). 

The Process Explained—Roughly

To give you a rough sketch of a smooth process, this is what hap-pens when you log on to the Net and type in a URL. The browser you’re using—Internet Explorer, Mozilla Firefox, Safari, Seamonkey,

or anything else—is called a client, which sends a “request” to

access the URL you typed in (say www.yahoo.co.in) to your ISP. The ISP

forwards the request to another (like a file on the table of a typical

government office goes to another table) through the vBNS (the

Very high-speed Backbone Network Services) and so on, until it

reaches the target where the destination is stored—in this case, the

Web site you requested. This “host” computer, or “Web host,” sends

back the packet of data you requested through much the same

channels in much the same way, hopping routers on the way, until

it reaches your ISP. The ISP then sends the data down to your com-puter and the Yahoo! India homepage appears on your screen.

Simple? No. Actually, mucho, muchocomplicated. 

 The Process Explained—Toughly 

We’ll now try to figure out how the Web works by digging a little

deeper and see what exactly happens to get your favourite Web

page on your screen.

Let’s say you typed in the URL of a Web site. The Uniform

Resource Locator, for your convenience, is in letters and numbers,

in recognisable words, so that you can identify and recall it at will.

The computers aren’t human… yet. So, out of our deep sympathy

for the dumb machines, we humans have conceded to talk to com-puters in their language, which is nothing but the language of

numbers. Computers converse with each other in the language of

zeroes and ones—“bits.” If all computers used this binary language

in random ways, exchange of meaningful data would be impossi-ble. To ensure a shared language between two computing

machines, we created “protocols.”

Protocols

A protocol is a set of rules that enables the exchange of information

between two computers, regardless of whether they run on differ-ent operating systems. These rules strictly state the format and pro-cedure to be followed while transmitting data between two com-puters or two networks. A protocol basically gives the “grammar” to their “speech.” Certain key protocols you should know about are:

TCP:Transmission Control Protocol, used for the reliable trans-mission of data over a network. 

IP: The Internet protocol is the set of rules that dictates how

exactly packets of data should be transmitted over the Internet.IP

standardises the way machines over the Internet (or any IP net-work) forward or route their packets based on their IP addresses.

HTTP:Hypertext Transfer Protocol, used for transmitting and dis-playing information in the form of Web pages on browsers. This is

the language used when your Web browser talks to the Web server. 

FTP:File Transfer Protocol, used for file transfer (uploading and

downloading) over the Internet. Both FTP and HTTP are based on

TCP/IP. FTP is still often used as a means of downloading large files.

SMTP:Simple Mail Transfer Protocol, used for e-mail.

Ethernet: Used for data transmission over a LAN (Local Area

Network). 

Wi-Fi:The wireless version of Ethernet.

Of all these protocols, in order to understand how the Internet

works, what we really need to know is the IP and the TCP.

 TCP/IP

Every device, every computer, that is part of the Internet, is allo-cated an address, called an Internet address, or Internet Protocol

address (IP address). However, it isn’t like our home address or any-thing because, as we mentioned earlier, computers only under-stand numbers. So the IP address is a series of numbers.

The current protocol is called IPv4. Each IP address is actual-ly just four numbers each ranging from 0 to 255 (each of which

is called an “octet”) and separated by decimal points (called

dots). IP addresses are, therefore in the format xxx.xxx.xxx.xxx

where each xxx could be any number from 0 to 255. So an IP

address could look something like 161.184.138.36. This address

is essential during every exchange of data on the Internet,

because it identifies the client computer (which makes the

request, and to which the data must be sent) and the destination

computer (to which the request must reach). Nothing, therefore,

can happen on the Internet unless your computer is first assigned an IP address.

Say you’ve specified the IP address of a Web site you wish to

access. A server usually has a static IP address. However, your home

PC, if you’re connecting to the Internet through a modem, is usu-ally assigned a different IP address by the ISP each time you dial

in. This IP address is unique to you as a user, but only for as long

as the session lasts. The reason for this system is that ISPs have to

deal with a large number of requests at the same time. Besides, at

any given time, the number of users of the ISP server is limited. To

assign permanent IP addresses to retail users wouldn’t really be

required; instead, ISPs simply allot any IP address that is not being

used at the time to each user who has just dialled in. This system

enables the ISP to get along with fewer IP addresses for the num-ber of users who are connected simultaneously.

TCP breaks down and reassembles packets of data. IP ensures

that the packets are sent to the right destination. The TCP/IP com-bination is used because the Internet is a “packet-switched net-work.” In a “circuit-switched network,” in contrast, once a connec-tion is made, that specific part of the network is exclusively used

only for that connection. In a packet-switched network, the con-nection between the sender and the receiver is not single and

unbroken. When information is sent, it is broken into small pack-ets and sent over many different routes at the same time and then

put together in order at the other end once the packets of data

reach the destination. 

How TCP/IP works

Data sent across the Internet is broken up into packets of less than

about 1,500 characters each. Each packet is given a “header,”

which holds the information needed to put the packets back in

right order again. To each header is adder a “checksum”—basical-ly a number that checks whether the precise amount of data in

the packet has been received or not. Each packet is put into what

is called an IP “envelope”—a packet of data with a common address

and other specifications. Once the packets go through the routers

to their destination, TCP, at the receiver’s end, calculates a check-sum for each packet and then compares it with the checksum that was sent in the packet. If these don’t match, the computer figures

that the data has been corrupted en route, discards the packet,

and requests the sender to re-send the packet. TCP then assembles

them into their original coherent form to allow the receiving

computer to make sense of it.

None of this would be possible, however, without a DNS server

acting as mediator.

 The Domain Name System

Most of us would find it considerably difficult to remember hun-dreds of strings of numbers of all the sites we visit. So, for humans,

the address is in letters (the URL). The domain name is the root

identifier of a Web site on the Internet.

When you want to contact a location on the Internet—say a Web

site—you’d type in an address like www.yahoo.com. IP uses (a) the

Internet address information and (b) the Domain Name System to

understand and deliver your mail or request from one computer to

another: the letters www.yahoo.comare translated into an IP address,

which is, in this case, 87.248.113.14. In fact, if you typed in

http://87.248.113.14, your screen would show you the Yahoo! site, and

you would make the job a microsecond faster and a trifle easier for

the computers. This translation is accomplished by the DNS server.

Before we understand what the DNS server does, let’s get a grip

of how the DNS works.

 How The Domain Name System Works

In order to make efficient and hassle-free the translation from the

plain-English www.yahoo.com into numbers that computers can

understand, the Internet has been organised into a number of

major domains. Major domains are represented by the letters at

the end of the English address. These are called top-level domains

(TLD) or zones, and they are either two or three letters long. The

three-letter zones indicate the type of organisation that owns the

domain, that is, whether it’s a commercial institution, ISP, non-governmental organisation of academic establishment, etc. All

Internet domains are registered with private companies called

Internet Registrars, in co-ordination with the InterNIC (InternetNetwork Information Center). You probably are most familiar with

the .com TLD. Here’s a list:

In October 1998, the Internet Corporation for Assigned Names

and Numbers (ICANN) was formed, and took over the task of managing the domain names and other responsibilities that were ful-filled by the InterNIC. ICANN announced on November 16, 2000,

the following seven new gTLDs:

Besides, additional two-letter domain names have been

assigned to identify domains belonging to particular countries (outside the US)—.au, .in and so on.So that was it .Awaiting for your precious reply