Unit 2

Communication
What is a WAN?
• 1970s
• Large organisations to have a single computer with TIME SHARING OS allowing individual users access
through terminals connected through cables. Such large organisations would be connected with other
organisations through telephone infrastructure forming a Wide Area Network.
• Today
• Organisations (like banks) connected various branches or sites with the help of a dedicated NETWORK
connected through fibre optics. The communication between these branches is handled by switches. Sites
have access to WAN but NOT the individual users.
Why use WANs?
• For running JOBs on a remote computer.
• For DATA STORAGE AND BACK UPs on a remote computer.
• For sending messages to a remote computer.
What is a LAN?
• 1980s
• Personal Computers allowed users to have their separate computer at their own desk. But to be able to
share some resources* they needed to be connected to each other. This gave rise to what we know as a
Local Area Network (LANs are PCs connected to each other within a single site or building)
• Today
• Used to connect (wirelessly or wired) a site or branch while forming part of a larger WAN.
• Each LAN is connected to another LAN using special hardware.
*Why use a LAN?
• LANs can allow for shared resources (and reducing costs) to individual users in a company such as :-
• Application being installed on an APPLICATION SERVER
• Large files being saved on a FILE SERVER.
• A common printer for an entire floor in the form of a PRINT SERVER.
• Mail server, allowing for “paper less” communications.
The Internet:-
• It is a network of networks (LAN(s)àWAN)
• With the rise of mobile devices and wireless
networks the scope of the Internet has enormously
grown.
The Client Server Model/Architecture:-
• The client (end system or web browser) is connected to a
powerful central computer or server (which can be part
of a LAN or a hosted on the Internet).
• The server will provide services to clients (for example by
allowing access to database for e-commerce or financial
transactions).

Peer to Peer (P2P) network:-

An alternate approach to C-S model for file sharing is a Peer to Peer
(P2P) network.
• In P2P network, a peer (a PC connected to the network) can act as
both a CLIENT and SERVER. (can request and provide data)
• If there is a single source (Client – Server) for services, many
connected clients may result in slow downloads. But in P2P
approach there are many “sources” for different parts of the same
file allowing for faster downloads.
• However, client server model has more control of the data being
shared therefore reducing the risk of malware being shared.
 The “Thin Client” vs “Thick Client”:-
• Thin Client a.k.a Dumb Terminals.
• They are compact/cheap computing that runs from
resources stored on server rather than local HD.
Thin Clients Thick Clients
• Users access the system through a VIRTUAL DESKTOP
• Easy / Cheap to install • Difficult / Expensive to install.
• Processing carried out by the server. • Results in more downtime* • More uptime*
• Depending on the performance/load of the server, the users • Requires continuous • Only needs intermittent
of a Thin Client may feel “lagged” services as compared to communication with server. communication with server
• Portable therefore reducing server
Thick Clients. • Makes good use of outdated demands.
• With improved capabilities, the difference between Thick PCs. • Can store local files and
and Thin clients has drastically reduced. • Reduced security threat applications

• The Thick a.k.a Fat Client

• Can carry out processing on it’s own.
*Downtime à the time during which the computer is
• If an application is being used on the server this approach can be unavailable for use.
handled in one of two ways:- Uptime à Duration when the computer is
• The data can be processed locally before and after it has been operational.
accessed from server.
• Or the the application can be installed and used locally thereby
reducing the time communicating with the server even more.
• Even though making using use of Thick Clients is expensive, they result
in lesser bandwidth requirements. Also, there is no need for continuous
communication with the server.
• Multimedia applications are bandwidth intensive therefore it is suitable
to go for Thick Clients.
Network Topologies:- Data can be transmitted using the following modes:-
For a Data Communication System you need to have the following:-
• Simplex :- Data is sent in one direction only.
• Sender • Half Duplex (HDX):- Data can be sent in both directions but not at
• Receiver the same time.
• Transmission Medium (Wi-Fi or Ethernet Cables) • Full Duplex (FDX):- Data can be sent in both directions
• Message (the data) simultaneously.
• Protocols (will be discussed later).
The massage or data can sent as either:-
• a broadcast message (one to all)
An isolated network may take shape in the following topologies (a • a multicast message (one to many)
configuration that defines how different devices in a network are connected) • a unicast message (one to one)
Point to Point Network:-
• It is the simplest configuration as only two systems are connected
together via dedicated link.
• Transmission can be simplex or duplex.
• Message is Unicast.
Bus Topology:-
• Early LANs used either Bus Topology or Ring Topology (rarely used
now will not be discussed in this course).
• Has one link which is shared by several end systems. i.e. Multi-
Point Connection.
• There is no direct link b/w end systems (computer or server
connected to the network) and messages are broadcasted (even if
they are intended for one).
• Duplex.
• It is a resilient configuration as any fault in an end system has no
impact functionality of network (provided that the central link
doesn’t get damaged).
Mesh Topology:-
• There is a point to point (direct) link b/w all end systems.
• Supports duplex communications.
• Messages can be unicast, multicast and broadcast.
• This is the configuration used for interconnected routers that are part of the
Internet.
• Using this configuration for connecting end systems is unrealistic due to cost of
cabling. However, individual LANs can be connected to a network through
switches.
Star Topology:-
• There is a point-to-point connection with each end system and a central device
(switch, hub or router).
• Transmission is duplex.
• Messages from central device may be unicast, multicast, broadcast.
• It is also resilient as failure of any end system has no effect over the over all
network. (provided the central device doesn’t fail).
• The central device can also be used to connect one network to another network
(even the Internet).
• The most preferred configuration.
Hybrid Network:-
• Various interconnecting LANs with different topologies or supporting
technologies form a Hybrid Network.
• Is beneficial in connecting a new topology LAN (such as wireless LAN) to an
existing (wired) LAN.
 Transmission Media Wireless
When using the term “transmission media” we are either referring to Factors to consider:-
cable (twisted pair, coaxial or fibre optic) or wireless (radio waves, infrared or • Cost
microwave). • Bandwidth
Cable • Chances of interference
Factors to consider:- • Attenuation (weakening)
• Cost • Penetration a solid barrier.
• Bandwidth • Directional focusing .
• Chances of interference
• Attenuation (weakening)

Repeaters vs Amplifiers
Repeater regenerates the original signal and then retransmits the regenerated signal. It will also
eliminate signal interference while reproducing signal bit by bit. Repeaters are used in fixed
environments such as buildings.
In contract Amplifiers amplifies a signal by increasing its “amplitude”. Furthermore, amplifier
simply increases the power of the entire signal without distinguishing b/w actual data and the
noise it carries. Amplifiers are used in mobile environments such as remote areas.
Twisted Pair
• Copper Transmission medium
• Currently used for telephone headsets and lines as well as ethernet LAN cable.
• Can be shielded (confined in foil or mesh shield to protect against electromagnetic interference) or unshielded (cables
are just twisted together to reduce noise).
Fibre Optic
• Best performance but highest cost.
• A bundled Fibre Optic contains many individual fibres (either glass or plastic).
• Data is transferred via rapid pulses (according to binary value) of light.
• Lesser noise but are more fragile therefore require more shielding. Coaxial
• Provide higher bandwidth and is the technology of choice long distance cabling. Which is • Copper Transmission Medium
why they are considered as a large part of the internet backbone. • Extensively used for cable television companies.
• Not used for long distance telephone cabling.
Infrared:-
• Data transfer is carried out using Infrared light waves.
• Has the highest frequency.
• Highest possibility to attenuation due to which its use is limited indoors.
• Cannot penetrate through walls.
• Mobile devices (PC or smartphone) and peripherals (mouse or keyboard) come equipped with an IrDA
(infra red data association) port for infrared data transfer.
Radio:-
• Radio (such as Broadcast Radio) have the greatest ability to penetrate solid
• Can be used for both transferring signals over long distance (Cellular Radio) as well as short distances
such as within an office or home (Bluetooth or WiFi).
Cellular Radio is an example of BROADCAST radio that is used for mobile communications (such as wireless modems
and cell phones) to transmit voice and digital data.
Microwaves a.k.a Fixed Wireless:-
• Provide high speed transmission between an earth-based station (GPS device or reflective dish) and space station
(satellite) that rely on “line of sight”.
• For best uplink (Earth to Space) and downlink (Space to Earth) transmissions, microwave stations must be
placed on top of buildings to avoid possible obstacles.
• Microwaves are used where installing physical media is difficult or impossible but line of sight is available.

Wired Transmission Medium Wireless Transmission Medium

• Permission granted by land owners for installation. • Due to frequencies, permissions is from government.
• Fibre optics are required for global communications. • Satellite transmissions work in tandem with fibre
• Repeaters are needed more often. optics.
• In small locations, wired and wireless are equally • Interference is more significant.
efficient. However, the cost and effect of cabling can • Are primarily used for mobile (cell) phones.
be a deciding factor.
The Role Of Satellites:-
There are three types of Satellites active today:-
• GEO
• Three GEO (Geostationary a.k.a Geosynchronous Earth Orbit) satellites at 35786
kms from the earth surface along the Earth’s equator are needed for global
coverage and long distance telephone/network communication.
• Geostationary means that the satellite orbits the Earth at the same speed as the
Earth spins. Therefore, from one point the satellite always appears to be at the
same point in the sky.
• MEO
• Ten MEO (Medium Earth Orbit) satellites (at more than 15000 kms from the
Earth’s surface) are needed for GPS (global positioning system) connectivity. NASA’s Syncom III was
• MEO satellites can transmit data at up to 1.6 Gbit/s, which is a much snappier the world’s first GEO
connection than most of us achieve through fibre connections to our homes. These satellite.
types of satellites are also used for functions such as GPS, GLONASS (GLObal Provided worldwide
NAvigation Satellite System) and Galileo television coverage of the
• LEO Olympic Games in
• Only fifty LEO (low Earth Orbit) satellites (max alt 5000km from Earth) are Tokyo
needed to supplement mobile phone networks. However, there several hundreds of
them up there!
• The Hubble Space Telescope, ISS, Iridium Satellites.
Because of their height they provide excellent Right now, the applications of satellites are specialised such as GPS or Internet
connectivity for networks that are geographically far Services in remote areas.
apart. However, greater distances cause transmission It was once believed that satellites could take over the bulk of internet
delays that end up causing problems for underlying communications the implementation of high-speed fibre optics at lower cost
networks. reduced their need.
 LAN Hardware
Wired LANs:-
• Earlier years COAXIAL cables were used for LANs. Now Twisted Pair are the most widely used. Fibre Optic cables are becoming more
common.
• The BUS configurations will have the following features:-
• The central (thick backbone main cable) bus will have a
series of sockets (T-Connectors) with cables (thinner branch
cables for end systems) attached to them.
• The ends of bus will have terminators* to prevent the signals
from reflecting down the bus.
• The branch cables have RJ-45 connectors on each end. One
is connected to the T-Connector the other the port on the
end system. BNC “Bayonet
• A bus can be extended by linking two buses together and Neill–Concelman”

using a repeater (due to attenuation over long distances).

• BUS network may be constructed in the form of segments
and each segment will be connected by a BRIDGE. The
bridge stores network address or MAC addresses of the end
systems in the two segments is connects in what is known as
“bridging table”.
• The LAN port on an end system is part of the Network
Interface Card (NIC) that is uniquely identified by its MAC
(Media Access Control) address. (to be discussed later in this
course).
*Terminators
Data can be sent in one direction at a time. As soon as the date reaches the extreme
end the terminator removes the data form the line avoid further collisions. Shown here
is an RJ 45 Terminator and RJ 45 Connectors for cables. à
A network bridge is a computer networking device that
creates a single aggregate network from multiple
communication networks or network segments. This
function is called network bridging.
• The STAR configurations will have the following features:-
• Each end system has the same type of cable with same connectors.
• The cable needs to be longer than in BUS as each end system must be
able to connect to a central device.
• This central device may be hub, switch or router. The switch being the
most common (as it can direct its unicast communication to a specific
end system).

• Wireless LANS
• WIFI (or WLAN some countries) is the term associated to a
wireless ETHERNET.
• The standard uses radio frequencies for data transmissions.
• The central device in WIFI LAN is a Wireless Access Point
(WAP).
• The WAP can communicate with an end system in the WLAN
provided that the end system has a WIRELESS NETWORK
INTERFACE CARD (WNIC).

• What is the difference between WAP and Router?

• WAP connects Wireless devices to an existing Wired LAN.
• Router performs two tasks. It acts like a traditional WAP as
well as a gateway between the LAN and a public network
i.e. the Internet.
Ethernet:- CSMA/CD:-
• Is a prevalent technology in the modern networked • Whenever a shared medium is used for data transfer there is a possibility of
world that is primarily focused on LANs. messages to be corrupted during transmission.
• The Ethernet standard (a.k.a IEEE 802.3) has been • This especially happens when we have two end systems that transmit at the
since evolved through 5 generations. (standard, fast, same time creating what is known as a “collision” (voltages of individual
gigabit, 10 gigabit and 100 gigabit). The “gigabits” transmissions interfere with each other thereby corrupting messages).
represent data transfer rates. • To deal with this the CSMA/CD method has been adopted which largely relies
• Original (or legacy) ethernet configurations were on the generation of a voltage level when the message was being transmitted on
either BUS topology OR Star Topology with a HUB the Ethernet cable that can be detected.
as a central device. Which meant that messages sent • The steps are as follows:-
were BROADCASTED. 1. Check the voltage on the transmission medium.
• Since each end system would receive all messages, 2. If this indicates activity, wait a random time before checking again.
every end system would have to check the destination 3. If no activity is detected, start transmission.
address (MAC Address) of incoming message to see if 4. Continuously check for a collision.
it was the intended recipient. 5. If no collision is detected, continue transmission.
6. If a collision is detected, stop transmission of the message and transmit a
jamming signal to warn all end-stations; after a random time, try again.

There may still be some traditional LANs operating now. However, most LANs use a SWITCH now. In a STAR CONFIGURATION with a
SWITCH as the central device each end system is connected to the central device via full duplex connection. Since the link is dedicated there
are no chances of collisions and that is why CSMA/CD is not needed anymore. Because there might be high levels of activity the switch
needs to be able to store an incoming message in a buffer until the cable is free for the transmission to take place
The Internet Infrastructure:-
Calling the Internet another WAN give us very little idea of its complexity. The Internet is the biggest internetwork in existence. Moreover, it
operates as a WHOLE (one entity). Interestingly, The Internet is still evolving.
 Internet Service Providers:-
Due to it’s immense size and constantly evolving state, one of the drawbacks of the Internet is
that there is no defined structure. However, if there some implementation of Hierarchy, it is in
the form of Internet Service Providers (ISPs).
By simple logic, an ISP is a company that provides internet access. The most common ISPs are
the ones that deliver Internet connectivity to your home. In reality, ISPs operate in three Tiers.
Tier 1 (Backbone) Tier 2 (Middle Tier) and Tier 3 (Access Tier). Connections between ISPs are
handled by Internet Exchange Points or IXPs (location where different ISPs meet to exchange
local traffic via a switch)
Tier 1:-
This tier exchanges traffic with other tier1’s through what is known as “peering
agreements” (backbones agree to exchange content amongst each other thereby allowing
global coverage).
These are more commonly known as the Backbones. They are responsible for
contributing to internet infrastructures such as the Internet sea cables (under water
fibres). Tier 1 ISPs provide traffic to other Tier ISPs and NOT to end users.
Without them Internet traffic cannot be exchanged between continents and countries.
Examples:- Hibernia Networks, Cogent Communications, Level 3 Communications
Tier 2:-
Tier 3:-
Tier 2 ISPs primarily connect the Tier 3s with
Tier 3 ISPs solely rely on
the Tier 1s. They operate on peering agreements
purchasing internet transit and
as well as purchasing Internet Transit (data
delivering it to homes and
being carried from one network or provider to
businesses. Tier 3 operate strictly
another and being charged according to the
on a purchasing model.
amount of data).
Examples:- Verizon and Comcast.
Examples:- British Telekom, Vodafone, Easynet
Router:-
The Internet can also be seen as a series of connections that used to carry out most of the traffic. These connections comprise of under
sea fibre optics connected via nodes forming a “mesh” structure. At these nodes is a device known as a “Router”.
These routers not only provide the general mesh for Internet communications but also for connectivity within an ISP network.
Each router is connected to several other routers and its function (apart from connectivity) is to also find the best possible route for a
transmission. More details of the working of a router will be discussed later in this course.
PSTN (Public Switched Telephone Network):-
- Is the basis for the original Internet infrastructure.
- In early stages of networking the telephone network only provided transmission of analogue voice data.
- For digital data to be communicated between computers both sender side and receiving side were equipped with “modems”
(modulator/demodulator).
- The internet services were provided via a “dial up network” that provided modest speeds to end users. For organisations, a leased line
was provided for higher speeds.
Cell Phone Network:-
There is an alternate method for Internet Access for devices with cell phone capability. Mobile Phone companies act as ISPs and cellular
devices connect to a cell tower to access the wireless telephone network that in turn connects the device to the internet.
Applications that make use of the Internet
What is the World Wide Web?
The Internet and WWW are often considered the same (huge misconception). The Internet is a Network of Networks (Internetwork). Whereas
the WWW is a distributed application that runs on the Internet which connects millions of webpages with each other by extensively using
hyper linking.
 Cloud Computing
An alternate approach is a PUBLIC CLOUD which is created, owned and managed by a third-
Cloud Computing is the process of providing
party cloud service provider. (Amazon, Google, Microsoft).
computing services to an end user via the
1. Is accessible using your browser making it easily accessible from anywhere using any device by
Internet.
any individual person or organisation.
PRIVATE CLOUD:- 2. Implemented on very large mainframe computers or server farms. They have massive amounts
1. Created, maintained and installed of available space allowing for easy scalability.
on-site only. Access is granted
through private network. The main advangtage of using Public Cloud is its versatility and “pay as you go” structure that
2. Installation in on-site but creation allows customers the provision of more capacity on demand. However, one disadvantage related to
and maintenance is outsourced to the use of a public cloud system is it’s lack of privacy. The cloud service provider has complete
a third party. access to all data stored on the cloud. This means that the user cannot be sure if their data is
3. Outsourced and accessible via the hidden from third parties. Moreover, the user is stricly relying on the cloud service provider with
Internet (not on-site). Access to respected to the security of their data.
only authorized users. The services provided by the cloud can be categorised in the following manner:-
A Private Cloud usually exists behind a • Infrastructure Provision (IAAS)
firewall and offer better security and • Platform Provision (PAAS)
control but are expensive. • Software Provision (SAAS)

Infrastructure as a Service Platform as a Service Software as a Service

Pay as you go for services Ready to use development Software that is made
such as storage, networking tools made available over the available for use without
and virtualisation. Internet. local installation. Accessed
through web browser.
Amazon Web Services Windows Azure Office 365
What is Bit Streaming?
Data is stored in bytes, but when compressed each byte may be compressed
into fewer bits. Compression is essential for data streaming therefore, it is
referred to as Bit Streaming as opposed to Byte Streaming. Streaming
Media can be of the following types:-
• The Source is a Website and the user can “download” the media file
and access it at a later convenient time.
• The Source is a Website but the user may not want to wait for the file
to be downloaded and can avail the “On Demand” Streaming option
• This a two-step process 1) the media is “buffered” on to a
location on the user’s computer 2) a suitable media playback SW
will then play back this file from the buffer.
• The content may also be created as it is being delivered to users. This
“live streaming” is best used for applications such as telecasting a
sporting event.
• There can be a large number of users, that is why the content
should be transmitted to a many content servers where can share
the burden on provided content to individual users.
For Media streaming a broadband link of at least 2.5Mbps is essential for
good quality movie presentation. However, such bit rates are not always
available therefore servers should hold videos with varying compression
rates to accommodate different user requirements.
IP Addressing:-
The source and destination for data transfer need to be defined. That is why we make use of IP addressing. Currently we are making use of IPv4 (four
bytes for each address each byte separated by “.” for example 128.12.2.30 (each byte is bound by limits previously discussed 0-25510 or 0-FF16 )
• 4 bytes à 232 à 4 Billion IP address (approx)
• each device connected with the Internet requires an IP address, therefore it is apparent that these IP addresses are not enough.
• Traditionally IPv4 is divided into two parts Net ID (bits used for identifying a network) and Host ID (identifying the connected device on that
network). Originally these IP addresses were divided in to three sub-classes.
• Class A: 0 <7bits for Net ID> <24 Bits for Host ID>.
• Ideal for Large Organisations (or networks) that could each host 224 different host. But there were only 27 (i.e. 128) such organisations
allowed.
• Class B: 10 <14 bits for NetID><16 bits for Host ID>
• Still only allowed 214 net IDs for networks.
• Class C: 110 <21 Bits for NetID><8 Bits for Host ID>
• Allowed for 221 different net IDs for network but each network could only host 28 different hosts (i.e. only 256 hosts which was too
small.
• IP addresses are too interpreted in binary (10000000 00001100 00000010 00011110) by underlying network HW. But for us to use these are
represented in equivalent denary (128.12.2.30) ß This is known as the dotted decimal notation.

Classless Inter-Domain Routing:-

• Is considered as the first approach towards improving the IP addressing scheme.
• It follows the same Net ID and Host ID pattern without the rigidity of any one class.
• The number of bits can be split into Net ID and Host ID according to individual need.
• This is done by adding a suffix at the end of the address to represent the number of bits (in binary) allocated for the Net ID.
• 195.12.6.14/21 à 110000110000110000000 11000001110 / 00010101
• Existing schemes (Class A, B and C) can also be used by using suffixes 8,16 and 24 respectively.
Sub Netting:-
This scheme makes use of only one Class C IP address. For example consider the following scenario:-
• An organisation has 7 different departments with with a total of 150 computers (hosts) each of which needs to be connected to the
internet. 8 different departments means 8 different networks. Which also means 8 different net IDs. Assuming Class C is being used, 8
class net IDs can each connect to 256 hosts. This results in 7 * 256 = 1792 hosts – 150(needed). = 1648 IP addresses wasted.
• Instead through sub netting we can use just ONE class netID and and organise the host ID byte in the following manner:-

001 00010
001 represent LAN 1 00010 represents
workstation 2

Using subnetting. We are making use of only ONE netID

with 256 hosts.
172.10.4.0 – 172.10.4.255.
• The NetID is unique, due to which Internet traffic is
routed to the Router for the organization’s network.
• Once a data packet is received by the main router, it
investigates the Host ID to identify the LAN (first three
bits).
• The packet is forwarded to the relevant LAN where the
router for that LAN investigates the remaining 5 bits to
identify the specific host and forwards the packet
accordingly.
Network Address Translation:-
• This scheme deviates from the basic rule of IP addressing where each IP address must be Unique.
• A single device such as a router acts as an agent between the Internet “Public Network” and a local (or “private”)
network. As a result of which only a single IP address is required to represent an entire group of computers.
• The scheme introduces two types of IP Addresses:-
• Public IP Address (issued by ISP):-
• Unique for Public Network
• Private IP address

• Is unique for the Private Network (Issued by Router).

IPv6:
There must soon be a migration to IP version 6 (1Pv6), which uses a 128-bit addressing scheme allowing 2128 different addresses, a huge number!
Ipv6 is represented using the COLON HEXADECIMAL notation in which the sequence is broken into 16-bit parts with each of these
represented by four hexadecimal characters. For example:-
68E6:7C48:FFFE:FFFF:3D20:1180:695A:FF01
In practice, this will allow more complex structuring of addresses. Documenting these addresses is not going to be fun.
1. Leading zeros in each 16-bit field are suppressed.
For example, 2001:0db8::0001 is rendered as 2001:db8::1.
2. Any “all-zero” field is written as 0.
2001:db8:0000:1:1:1:1:1 is rendered as 2001:db8:0:1:1:1:1:1.
3. Where we have consecutive 0 fields, "::" is used.
For example, 2001:db8:0:0:0:0:2:1 is shortened to 2001:db8::2:1,
4. If there are multiple longest runs of all-zero fields, then it is the leftmost that is compressed.
E.g., 2001:db8:0:0:1:0:0:1 is rendered as 2001:db8::1:0:0:1. Whereas 2001:db8:0:0:1::1 would be incorrect.
5. Hexadecimal digits are expressed as lower-case letters.
For example, 2001:db8::1 is preferred over 2001:DB8::1.
6. Representations are shortened as much as possible.
Domain Names:-
Domain names are easier for us to understand but impossible for Routers to understand. Therefore, Domains names must
be translated to IP address (which are difficult for us to use).

The Domain Name System is made up of the following:-

- ROOT SERVERS that contain a list of of all TOP LEVEL DOMAINS and their associated numeric addresses.
- There are more than 250 top level domains that are either generic (.com,.edu. and .gov) or represent countries (.pk, or .uk).
- There are 750 different copies of Root Servers around in the world in all six continents. They are accessible through 13 root server IP
addresses. Since there are only 13 root servers, only 13 can be seen from any single location at a time. Different servers (with the same IP
addresses) will be seen from different locations. Universal Resource Locator
https://sites.google.com
 Authoritative Vs Recursive DNS Servers
The queries that you make are responded by either an Authoritative Server or Recursive DNS Server. They hold the standard information
that maps IP addresses to domains.

Domain Name Resolution Process (looking up a domain name to find an IP address):-

Let’s consider that you entered the URL (uniform resource locator) sites.google.com. The domain name resolution will take place in the following
manner:-
• The browser will search it’s local cache (in the local disk) for it’s IP address. If found the address will be used to foreword message.
• If the IP address is not recovered from the local cache, the browser will request a Recursive server for the address. If recovered, this address will be
forwarded to the browser. The browser will use this address to forward the message. The address will also be saved in local cache.
• However, if the Recursive Server doesn’t have the address, then it will “find” the IP address using the following steps:-
• The recursive will request the ROOT SERVER for the address for “.com” Authoritative Server.
• Next the Recursive will contact the .com authoritative for the address to google.com authoritative.
• The recursive server will contact google.com authoritative for the address to sites.google.com.
• Once the address for sites.google.com is completed, the Recursive Server will save a copy and forward the address to your browser. The browse will
save a copy locally and use this address to forward message.