Module 3

Part 2
HISTORY OF CLOUD COMPUTING
1. Client-Server Technology
• Client-Server is the technology behind cloud computing
• software applications and data are kept in Server
• If a user requires a data he needs to connect to the server
• The client requests to the server,
• the server replies appropriately.
• The client is known as dumb terminal as it does not have high processing power, storage space, or
memory.
• The client is simply a tool for user facilitation.
• Clients have to wait for getting the server
• In a client server, instant access is not possible all the time.
• The client server also has a centralized storage
• Thin clients and fat clients
• Fat clients
• A computer in a client–server technology that is independent of the central
server. Fat clients make periodic connections with the server.It easily
performs computations, when compared to thin clients
• A thin client
• A computer that is used to create a virtual environment. It depends on the
server for performing computation. Desktop, files , storage can be shared The
advantage of using thin clients is to reduce software maintenance cost and
maintain security of data.
Advantages of Client–Server Networks
• Centralized control There is a centralized power used in the client-
server model. Servers controls the complete system. login of different
users and resource allotment is made by servers.
• Administration managed File administration becomes easy since
every file is saved at a single place.
• Replication for backup Since complete data is saved on the server, it
is easy to create a back-up of it.
• Easy updating possible Variations may be made simply by mere
advancement of the server.
• Remote service Access of distant server is possible to accomplish the
needs of consumers and support distinct network.
• Secure and safe. Safety convention and access privileges can be set at
the server.
Disadvantages of Client–Server
Networks
• If the server is unable to handle countless requests from the
consumers, it might lead to congestion and data will be lost.
• If a file is being downloaded and if it is discarded at the server due to
some fault, download also fails
• Administration and deployment is extremely costly
• Expert IT professionals are required for maintaining servers and other
technological particulars of the network.
2. Peer-to-Peer Approach
• Peer-to-Peer (P2P) is a decentralized approach
• No main server.
• Every computer has equal responsibilities and facilities.
• In a client server, one or more computers are devoted to serving the
others.
• In a P2P setting, there is no master (boss) and slave.
• P2P facilitates straight swap of services and resources.
• not simple to manage.
Disadvantages of Peer-to-Peer
Networks
• It is not easy to administer this type of network.
• Safety cannot be tackled appropriately.
• Backup or data revival is not simple. Every computer must have its
personal imitation system.
3. Distributed Computing
• To increase the throughput and for the maximum utilization of the
computing power of a system.
• There are idle computer systems
• if the idle time of various systems is integrated and used for
computation, it is known as distributed computing.
• Distributed computing utilizes the idle resources that are not utilized
• Thus there is highest usage of resources
Disadvantages of Distributed
Computing
• Implementation cost is high, security is less, more complex,
• Network errors can happen
• Under-consumption of the resources in various scattered data centers
• Rise in maintenance expenses of data centers
4. Evolution of Cloud Computing
from Grid Computing
• Grid computing utilizes a network or group of computers for making
computing resources like super computers
• computers may be at different locations.
• Unused computing powers of many computers are used to perform
complex scientific problems
• 2 types – tightly coupled and loosely coupled
• Tightly coupled systems – connected by a system bus, normally share
memory and other system resources, along with the processors.
• Loosely coupled- independent memory and processor. Systems
communicate only when needed
Advantages
• Unused computer resources are utilized
• It can satisfy the need of high computing power.
• The resources of various computers can be shared with some
understanding among the users. (no centralized management of
resources)
• servers are connected with each other in an organi
• there is one administrator to control the servers
• services are accessed by the clients on the network.
5. Autonomic Computing
• Computers perform and work without any outer control or
intervention.
• The aim of autonomic computing is to have the computer carry out
versatile and critical functions without any interference by a user.
6. Platform Virtualization
• Virtualization is a base for clouds
• In virtualization , an undefined number of hardware resources is
available.
• Virtualization hides physical resources and shows as a virtual
resources for users.
• Virtualization is managed by a Virtual Machine Monitor (VMM), called
a hypervisor.
• Eg: Xen, Vmware
7.Service Oriented Architecture—SOA

• Cloud computing is Internet-based computing

• Information, software, and resources are offered to consumers on-need.
Service-oriented-architecture (SOA) is a flexible, modular approach to delivering IT
services. It is an essential foundation for emerging technologies like cloud.it provides
increased flexibility to companies according to changes occurring.

A good example of a SOA-based system is a set of customer services, like CRM, ERP,

Product Information Management System (PIM), etc. These services can be implemented

using different technologies and support diverse protocols of communication, data

models, etc. However, to enable effective service-to-service communication and align

their workflows, the system needs an additional integration mechanism in between.

• service level agreements (SLAs) are compulsory to manage the usage of resources.
• A joint decision is made between the service provider, and service consumer .
• Service aggrement is made on data loss, delay and bandwidth

• Quality of service (QoS) is also ensured

8. Utility Computing
• Utility computing allocates resources as a metered service when
required.
• Offers computing services on-demand, pay-per-use system.
• the supplier possesses and controls the computing resources and
infrastructure
• This is accessed by subscribers as and when needed on a metered basis
• Utility computing may offer virtual software, virtual storage, virtual
servers
9. Web 2.0

• A change in technology in WWW

• It is designed to increase data security and customization of application
• Web-based service hosting, social networking, sharing videos, etc., are
important provisions provided by Web 2.0.
The important features of Web 2.0 are as follows:
• Easy to access
• User interaction and participation
• Rich customization features
• Easy communication through video chatting, instant messaging
facilities, etc.
• User-friendly writing tools and applications
• Data management and analysis
• Multimedia supporting tools
• Web application and hosting
10. Parallel Computing

• Basic principle is that a single large problem is divided into small parts
and parallelly runs different parts on different machines.
• Geographically dispersed computer resources are used for a common
purpose.
• Different parts of the same problem run in parallel.
• Working of parallel computing:
• 1. Discrete parts of a broken problem can be solved in parallel
• 2. every part is broken down into a series of instructions
• 3. Different processors are used to execute the instructions in parallel
• 4. Coordination method works behind parallel computing
• Parallel computing supports applications that require processing of a
large problem..eg are
• Big data problem
• Data mining
• Search engines
• Medical diagnosis
• Virtual reality
• Multimedia
BENEFITS OF CLOUD COMPUTING

• Pay as per use On demand access to pay-as-you-go computing resources

and ability to release these computing resources when they are no longer
needed.
• Reduced investment and proportional costs :cloud providers make IT
resources available to cloud consumers in attractive prices. So organizations
can gain access to powerful infrastructure without having to purchase it.
Expenditure on hardware and software is called proportional cost.
• Accessibility from anywhere Resources can be accessed from anywhere,
irrespective of location and device. So there is business continuity around
the clock.
• Increased scalability The business load can be handled with scalability of
the respective resources without much effort, time, and cost.
• Increased availability and reliability Resources ( services )are
available all the time. Even if there is a failure ,the provider
immediately identifies, isolates, and replaces the failed components
without any performance degradations.
• Dynamic provisioning : Unlimited computing resources are available
on demand without any provisioning time.
LIMITATIONS OF CLOUD COMPUTING

• Availability of Services
• Data Lock-in
• Data Segregation
• Privilege Neglect
• Scaling Resources
• Data Location
• Deletion of Data
• Recovery and Backup
• Offline Clouds
• Unpredictable Performance
• Availability of Services
Other than conventional applications, there are more ways to access the
application or information over cloud computing. Consumer may sometimes
discard the data or sometimes recover the whole data.
• Data Lock-in
shifting of data and applications from one platform to another is a challenge to
the cloud provider . Google has a scheme known as Data Liberation Front, to
support user shifting applications and data in and out of their platform.
• Data Segregation
It is difficult to isolate cloud users from each other. virtual machines of distinct
consumers are co-located on a single server or data on single hard disks, is a
privacy issue.
• Privilege Neglect
Sensitive data of a company can be disclosed by a malicious insider who has
access to confidential data.
• Scaling Resources
Scaling up and scaling down of resources is chief advantage of
Cloud computing. In a multi-tenant environment (multitenancy), a
single storage device is shared by more than one cloud user In the
figure, there are two consumers 1 and 2, who are sharing a single
shared storage for storing data, so there is the risk of interchanging
or risk related to mismatch of data if proper arrangement is not carried
out.
• Data Location
The geographical site of the data is also important for companies. Data should
be stored at the smallest distance from users. Some cloud providers give the
users a choice to select the location(eg: Amazon)
• Deletion of Data
Sometimes a company wants to completely erase its data from cloud or wants
to migrate to another cloud provider. If the deletion is not complete, the data
will be used by a malicious user.
• Recovery and Backup
keeping the data of a consumer safe at different locations is abig
challenge. There should be an agreement between users and cloud
providers on this backup and recovery.
• Offline Clouds
A business on cloud computing always depend upon Internet
availability and this is not always feasible. So there should be offline
methods to access the data in the cloud
• Google launched Gears, a free-of-cost add-on for the browser, which allows
data to be saved locally.
• Latest HTML protocol tackles the offline matters with a pair of constituents—
database and AppCache.
• Unpredictable Performance
The end user cannot estimate the performance provided by cloud companies.
The performance depends on several aspects and the end user has no control
over many of them.
Cloud Infrastructure

• Infrastructure should meet the following vital features

• Accessibility It is the major responsibility of every central data store
that data is accessible to users.
• Functioning Optimal service and functioning should be offered by the
data center.
• Supple High-priority business needs should be satisfied without any
disruption of accessibility and with the least charge.
• Safety Data must be protected from illegal and malevolent users.
• Manageability :Reduced IT expenses on the data center and increase
the consumption of prevailing resources.
• For the development of cloud infrastructure, the following are needed:
1.Understanding the present conventional data center
2. understanding Computing resources that will be virtualized
3. Installing service administration devices
• Cloud infrastructure may be built in several stages:
1. thoroughly recognizing the prevailing physical infrastructure
2. providing the prevailing infrastructure resources by using
virtualization technologies.
3. install service administration devices to reduce human
interference. Service administration devices include services and
utilization-based metering so that customers have to pay for just what
is utilized by them.
Vendors in Cloud computing
• Amazon Web Services (AWS)
• Offers Iaas
• Elastic Compute Cloud(EC2)- user can create virtual machines (Amazon
Machine Image AMI), pay-by-the- hour system
• Simple Storage Service (S3)- regain and store infinite amount of data anytime
anywhere
• Google
• SaaS and PaaS
• Google App Engine –PaaS- Python, Java- payment for storage, bandwidth, pu
cycles
• SaaS- gmail, calendar, google docs, google sheet
• Microsoft Azure (PaaS)
• Rackspace-cloud hosting
• Salesforce.com- PaaS, SaaS