See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/314072571

Cloud Computing - A Comprehensive Definiton

Article · January 2017

CITATIONS READS

5 6,765

1 author:

Ishrat Ahmad
Cardiff Metropolitan University
2 PUBLICATIONS   10 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Cloud Security View project

All content following this page was uploaded by Ishrat Ahmad on 20 September 2017.

The user has requested enhancement of the downloaded file.

 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

Cloud Computing – A Comprehensive Definition

I. Ahmad, H. Bakht, U. Mohan

Abstract—Cloud computing is an evolving technology that is consistently generating impact in IT industry and academia. It performs all
computational tasks over Internet by using virtualization techniques and remains isolated from intricate vast hardware and software infrastruc-
tures. The aim of this paper is to glance through the background and evolutions of cloud computing, its architecture and services to develop
cumulative knowledge for future research extension and evaluation.

Keywords: cloud characteristics, cloud computing, cloud deployment models, cloud services

——————————  ——————————

1 INTRODUCTION

C loud computing is becoming dominant day by day.

It is the most happening trend that makes companies
to use the term in their marketing campaign in order
novation and product development. Cloud based busi-
ness applications is an inexpensive approach. Cloud serv-
ice providers bill for each instance, like pay-as-you-go, so
to draw people more into their products. Companies use better accountability and transparency of costs [5]. A
their data in so many ways to meet business needs, and cloud user is being charged based on the subscription
so, computing technologies must be able to process and model, i.e, one pay for what one uses, hence cost trans-
store the on-demand business data consistently and effec- parency and better accountability. It is scalable and con-
tively. Since traditional distributed computing architec- sumes less time to start up. As an example, New York
ture lacks in delivering flexibility, scalability, elasticity Times converted 11 million articles to portable document
and faster processing speed, companies are diverging into files (pdf) by renting Amazon E2C servers for a day at a
cloud technologies for huge data storage and load- much lower cost, that would have required weeks or even
balancing availability [1]. months to complete the task [6].
Cloud computing is basically a philosophy and design Many studies [1-6, 15, 21, 29-32, 40-43] have been con-
concept of computing architecture, while, its much more ducted on cloud computing from different perspectives,
complicated but yet so much simpler. The basic concept is such as, definition, technology, architecture, services and
to separate the Applications, Operating System and deployment models. Apparently, none of the study has
Hardware from each other. In case of any failure or a vi- covered all perspectives together in one particular study.
rus attack, for example in an operating system, instead of For example, in terms of cloud characteristics, which are
shutting down the whole system, the application can just eight in nature; some studies have covered only six, five,
be migrated automatically to another server by using vir- four, or even two, respectively [29, 21, 31, 30, 2]. Another
tualization technology. Each physical server can host sev- example in terms of deployement models, some studies
eral virtual servers and each cloud user can have one or have discussed private, public, community and hybrid
more virtual instances to store or host their data on the cloud [3,17, 24, 29, 31, 36], while another study has intro-
cloud servers. A study by IEEE members, Youseff, Butrico duced virtual private cloud [21]. This paper covers all the
and Silva [2], view cloud computing as a re- possible defintions of cloud characterisctics, architecture,
conceptualized and advancement in technology rather deployment models and services. The aim is to provide a
than disruptive innovation and they also view as the as- complete overview of how cloud entities work together
sembly of pre-existing technologies and components. Ac- with a solid ground of technical infrastructure, which will
cording to National Institute of Standards and Technol- enable any researchers to develop cumulative knowledge
ogy (NIST), cloud computing can be implemented in dis- for future research extension and evaluation.
tinctive architectural forms by using existing service and
deployment models, which is integrated with other tech-
2 CLOUD COMPUTING BACKGROUND
nical components and software tools [3]. The introduction
by Office of the Privacy Commissioner of Canada (OPC) IT industrialists have defined cloud computing from their
stated that cloud computing is a very inexpensive ap- respective business perspectives. US National Institute of
proach as it saves costs from expensive real-estate, elec- Standards and Technology (NIST) defined as, “Cloud
tricity cabling, cooling equipment, hardware, network- Computing is a model for enabling convenient, on-
bandwidth, operational costs, physical security, just to demand network access to a shared pool of configurable
name a few [4]. computing resources (e.g. networks, servers, storage, ap-
No human resources are needed for IT infrastructure plications, and services) that can be rapidly provisioned
maintenances as it provides outsourcing facilities. It helps and released with minimal management effort or service
organizations to rent services for a short period of time to provider interaction” [3]. According to Varia [7], technol-
accomplish a large task for their business purpose and ogy evangelist of Amazon Web Services, “Cloud Archi-
hence allows organization to focus more on business in- tectures are designs of software applications that use In-
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

ternet-accessible on-demand services. Applications built Platform as a Service and Software as a Service.
on Cloud Architectures are such that the underlying
computing infrastructure is used only when it is needed 2.1 Evolution of Cloud Computing Services
(for example to process a user request), draw the neces- Cloud computing is driven from modern technological
sary resources on-demand (like compute servers or stor- features in order to create a novel computing environ-
age), perform a specific job, then relinquish the unneeded ment that can serve various growing business require-
resources and often dispose themselves after the job is ment with greater flexibility, integrity and agility [15].
done. While in operation the application scales up or Various literatures relate cloud computing to prior para-
down elastically based on resource needs.” According to digms as Cluster Computing, Grid Computing, Virtual-
IBM, “The concept of cloud computing has developed ization, Utility Computing and Service Oriented Architec-
from earlier ideas such as grid and utility computing, and ture [2,16, 17, 18, 19, 20]. Others refer cloud computing as
aims to provide a completely Internet driven, dynamic a new paradigm and emerging technology [21]. Larry
and scalable service-oriented IT environment, which can Ellison, chief Executive of Oracle Corporation, said at a
be accessed from anywhere using any Web-capable de- financial analyst conference in 2008 that, “The interesting
vice” [8]. Microsoft defined as, “Cloud computing repre- thing about cloud computing is that we’ve redefined
sents the platform for the next generation of business. cloud computing to include everything we already do”
Cloud computing is driving the transformation of the IT [22]. The increased amount of data and high level of con-
industry across the entire stack: hardware model deliver- nectivity has made information technology providers to
ing incredibly powerful and efficient hardware at a frac- establish data centers that help for dynamic load balanc-
tion of the cost; application model allowing developers to ing, distributing and replicating data across servers on
rapidly create highly available secure cloud applications; demand [15].
operations model keeping cloud applications available Eric Schmidt, the executive chairman of Google, is
24X7 with 9-to-5 management” [9]. Sun views cloud as, “a probably the first person to introduce the concept of
set of services that are encapsulated, have an API, and are cloud computing “as a whole” in his talk at a conference
available over the network. This definition encompasses of Search Engine Strategies [23] cited by [24]. Prior to that,
using both compute and storage resources as services” the journey of cloud computing has started by Sales-
[10]. VMware finds cloud computing as, “is best under- force.com in 1999 with limited specific software packages
stood from the perspective of the consumer of services [22]. They later introduced integrated set of tools,
provided by such a computing approach. The following Force.com, which can be used by other organizations to
four attributes are core to cloud computing based serv- build their business applications and share the same in-
ices: radically improved economics through shared infra- frastructure that delivers the Salesforce CRM applications
structure; pricing based on consumption: You only pay [25]. From 2003 to 2006, Google has published some re-
for what you use; flexible access; a lightweight entry and search papers on Platform as a Service (PaaS) cloud com-
exit service acquisition model” [11]. Xiao and Xiao, senior puting and later launched its service to the public as
member of IEEE, defined cloud computing as, “A large- Google App Engine (GAE) in 2008 [24]. GAE provides
scale distributed computing paradigm that is driven by SaaS services as office productivity tools, including e-
economies of scale, in which a pool of abstracted, virtual- mail, calendar, and word processing and more in the area
ized, dynamically- scalable, managed computing power, of electronic corporate communications [25]. By the end
storage, platforms, and services are delivered on demand of 2013, Google has launched its on-demand Infrastruc-
to external customers over the Internet” [12]. They further ture as a Service as Google Compute Engine (GCE) and
mentioned that several literatures have given different distributed storage system as Bigtable. The GCE is hosted
definitions for cloud computing but none has been recog- on Google server and has an environment based on Py-
nized extensively. In another research by Voas and Zhang thon programming language. Cloud users access GCE
[13] stated that cloud computing is influencing significant platform to build and host web applications. During 2006
changes to the use of IT resources over Internet with a and 2007, Amazon has started offering a suite of several
standard interfaces. It is considered as a killer technology services under the name of Amazon Web Services (AWS)
that takes the best features from prior technologies and [24]. It includes cloud computing as Elastic Compute
trades them off to convert a new solution [14]. Cloud (EC2) [26] with no constraint on the programming
In general, a cloud comprises of memory, data centre, language type of software application [22], structure data
distributed storage, virtual machines, web applications, storage as SimpleDB [26], object storage service or Simple
processing and network. It has compatibility with loosely Storage Service (S3), a web service for content delivery as
coupled CPU clusters. According to the requirement of CloudFront, a hosted service for storing messages as
various organizational services, cloud can be deployed as Simple Queue Service [25] and more, in which, data stor-
any of the three main models, such as, Public clouds age are priced by the GB monthly and computing capac-
(available to general public), Private clouds (owned by ity by the CPU hour [28]. AWS is completely based on
one or groups of organizations) or combination of both as server virtualization technology and has made Google the
Hybrid clouds. The deployment can also be modeled as pioneer of Infrastructure as a Service (IaaS) provider [24].
Domain Specific clouds or Community clouds. The cloud At the end of 2008, Microsoft Azure was introduced to
architecture consists of abstract layers to provide three the public with services that includes BLOB object storage
different types of services as Infrastructure as a Service, and SQL services. It uses Window Azure Hypervisor
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

(WAH) as the core infrastructure with the capability of 3.4 Rapid Elasticity and Scalability:
server virtualization, and .Net as the application con- The scalability of the available resources appears to be
tainer with existing software and applications compatibil- unlimited at any time for better services. Nodes in the
ity [24]. In 2010, Windows Azure operating system PaaS network can be added or dropped with less alteration to
was launched to provide Azure services as Live, SQL, infrastructure setup. It is acquired and released automati-
SharePoint, .Net and Microsoft Dynamic CRM [25]. cally corresponding with the customers demand [29, 30,
In 2011, Apple released iCloud that act as an Internet 31].
repository. iCloud synchronizes all the data, files, appli-
cations and other items among user’s devices, like laptop, 3.5 Measured service or Utility-based pricing:
mobile etc. so that the devices can share the same data Resource usage can be monitored, controlled, metered
with no redundancy [27]. and reported in order to provide transparency for both
Since 2007 number of companies started walking the cloud service providers and users. The cloud services
through the journey. Cloud computing has become the are automatically measured at different level of abstrac-
key strategy of IT industries, ISP and telecom service pro- tion to the type of service layers (e.g., storage, memory,
viders. Following are the name of some cloud service servers, user applications, active accounts etc.) using ap-
providers: IBM Blue Cloud infrastructure, Joyent (Accel- propriate metering systems and are billed accordingly
erator), OpenNebula, Cisco UCS, Hadoop, Navajo SaaS, [29, 30]. The pricing scheme in cloud computing is like
GoGrid (Xen), Apache HBase, Verizon CaaS (VMware), pay-as-you-go and may vary from service to service. As
Heroku (Ruby), EMC Atmos, AT&T Synaptic (VMware) an example, a cloud customer can rent a virtual machine
[24,28]. Some of the key cloud computing technology on per-hour basis. Such utility based pricing model cuts
providers are Apache, Cisco and EMC. Example of serv- down service operating costs [21].
ice support providers are CapGemini, RigthScale and
Vordel [25].   3.6 Location Independence:
Usually, the cloud customer has no knowledge about the
exact geographical location of the rented resources from
3 CLOUD CHARACTERISTICS   service providers. In order to provide maximum service
Cloud computing has some key characteristics that depict utility and high network performances, many service
their features of similarity and differences from conven- providers set up their data centres at different regions
tional computing operations. The first five essential char- around the globe. And may specify the location to the
acteristics described below are defined by NIST [3] and customers only at higher level of abstraction, as in like,
CSA [29]. Other relevant studies have also identified a country or state [21, 29, 31].
number of salient features of cloud computing, and they
are summarized as follows:
3.7 Cost Effectiveness:
The cloud can be deployed near to economical power
3.1 On-demand Self-service: station and in inexpensive real estate [31].
The cloud users have control over computing resources
that can be allocated or de-allocated and customized ac-
3.8 Multi-tenancy:
cording to their needs at anytime from anywhere without The cloud service provider can rent out a single infra-
any intervention from cloud service provider. They are structure as a service to multiple customers by providing
empowered to manage the resources and have adminis- appropriate technical partitions [21, 32] named multi-
trative privileges to configure the parameters of comput- tenancy and location independence as a Shared Infra-
ing power, servers, storage and network. Such automated structure characteristic and has defined that the cloud
resource management provides greater flexibility and fast customers use physical services as virtualized software
response to rapid changes on service demand [3, 21, 29]. model in accordance to their demand.

3.2 Broad Network Access:

4 CLOUD COMPUTING ARCHITECTURE
The cloud users can access the services over the network
or Internet through standard interfaces of heterogeneous Cloud computing is a promising revolutionary of the con-
thin or thick client platforms, e.g. mobile phones, PDAs, temporary distributed computing resources. These re-
tablets and laptops [29, 30, 31]. sources are accessible as services by cloud users from
cloud service providers. According to different business
3.3 Resource Pooling or Provisioning: needs, cloud architecture can be categories into four basic
The cloud service providers use multi-tenant model by types of deployment models and three types of service
pooling computing resources to serve multiple customers. models. Each service has distinctive fundamental charac-
Resources include virtual machines, storage, processing, teristics, with its own benefits and weaknesses. NIST and
memory and network bandwidth. The feature allows CSA have outlined the cloud architecture in [3, 29], which
these resources to dynamically assign or reassign accord- are discussed below.
ing to the customers demand that can be configured to
appear as a distinctive platform for individual customer
[3, 21, 29].
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

4.1 Deployment Models attack or leak of information, as the security level in pub-
lic cloud lacks fine-grained control mechanisms [3, 17, 29,
4.1.1 Private Cloud 31, 36].
The private cloud infrastructure is configured solely for a
single organization or a third party by using virtualiza- 4.1.3 Virtual Private Cloud
tion and automated management technologies to ensure Virtual Private cloud is one of the emerging cloud infra-
highest degree of controls over security, reliability and structures that have started to gain popularity since its
performance. It is built, deployed, customized, managed, establishment in 2009 [21]. The model is configured to run
operated and used by either a single or group of organi- on public cloud but in a private way. It uses the comput-
zation as client and supervised by cloud service provider. ing resources in public cloud and allows connection with
It is not made available to the general public [33]. Figure 1 consumer’s computing resources by using virtual private
shows an example of a Private Cloud model. network (VPN) technology. Hence, better security set-
tings options that can virtualize not only the servers and
applications but also the underlying communication net-
work [21]. Google and Amazon provide services as Vir-
tual Private cloud (VPC), which is a virtual platform op-
erating on public cloud and is shared by different organi-
zations [17].

4.1.4 Hybrid Cloud

The cloud infrastructure is a combination of two or more
Fig. 1. Private Cloud distinct cloud infrastructures (public, private or commu-
nity) and is considered as a single model. It is configured
The model may exist on or off premises and offers a new by standardized or proprietary technologies to share re-
architecture for improved efficiency and can be used to sources between clouds over a flexible controlled and
provide enterprise level computing services or B2B inter- secured network. The cloud users offload their applica-
actions, such as ERP applications, online collaboration, tion processes and data to acquire benefit of scalability
email, calendaring etc., unlike conventional servers that and cost effectiveness, e.g. cloud bursting for load balanc-
are dedicated only for certain applications. Customer can ing [29]. It has more flexibility over other cloud infrastruc-
maximize or optimize the resource utilization when tures by allowing highly confidential applications to run
needed; no data privacy concern and can always control in private cloud while for general peak loads or computa-
business critical activities within network firewalls [34]. tion in public cloud [37]. Figure 3 shows an example of a
The Aberdeen Group [35] has claimed in a report that Hybrid Cloud model.
organizations operating using private cloud are leading
by about 12% in cost than organizations operating in pub-
lic cloud [3, 17, 21, 29, 31].

4.1.2 Public Cloud

The cloud infrastructure is configured by the cloud serv-
ice providers and made available to general public or ex-
ternal parties to use over Internet. It is owned, managed,
operated by the cloud service provider and exist on their
premises. Figure 2 shows an example of a Public Cloud
model. Fig. 3. Hybrid Cloud

In other words, less critical information is compute

into the public cloud, while the business critical informa-
tion is stored within the private cloud [26]. Basically, B2B
or B2C interactions can be performed, by using portability
techniques over computing resources amongst different
clouds [37]. Despite facilitating on-demand service ex-
pansion and contraction, the partition between clouds
infrastructure need to be done very cautiously in order to
Fig. 2. Public Cloud
form hybrid cloud [21, 24]. The concept of the develop-
ment and interoperability of hybrid cloud is still at an
early stage and needs major consideration to overcome
The model offers flexibility and easy access of computing
any additional complexities over usual tools and tech-
services to general public or industry group without let-
nologies that are used to manage public or private cloud
ting them to invest huge capital on infrastructure. Despite
[3, 24, 31, 36].
of major benefits, public cloud has prone to get malicious
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

4.1.5 Community Cloud ready-made web-based or desktop applications are range

The cloud infrastructure is configured and shared by sev- from basic online email accounts like Hotmail or Gmail to
eral organizations that share similar domain of business, complex tools like CRM applications [41]. A SaaS model
for example, financial services, healthcare services, mis- is implemented to provide benefits of commercially li-
sion or security services, policy or compliance considera- censed enterprise level business application at a very low
tions. Figure 4 shows an example of a Community Cloud cost [42]. This model appears to be most appealing to
model. cloud customers because of the reduction of software and
system maintenance expenditures or upfront costs into
hardware and software licenses. Additionally, there is no
hassle for software upgrade as patches and features can
be deployed centrally [40]. The service provider manages
and controls the core cloud infrastructure including serv-
ers, storage, network, operating systems, applications and
even the security clauses for the customers’ information,
while, the customer only administers some limited appli-
cation configuration settings [30, 31, 38, 43]. A SaaS model
provides multi-tenant applications, which are shared, but
logically distinctive, to multiple customers [39].
Fig. 4. Community Cloud Examples of SaaS applications: Facebook, GmailTM,
Yahoo user applications, Google Docs, Microsoft online
services and more [37, 43]. Examples of SaaS providers
include Salesforce.com (provides online CRM), Windows
The model is owned, managed and operated by an orga- Live Mesh, Rackspace and SAP Business ByDesign, Goo-
nization or a third party and may exist on or off premises. gle Drive (online storage), SpringCM (content or docu-
The level of security is much higher than a public cloud ment management), Oracle (on-demand for CRM), Ning
and less than a private cloud [3, 29, 31]. (social networks), Zaora (billing), Google Apps or MS
Office (desktop productivity), StreetSmarts (sales), Xero
4.2 Cloud Services (financials) and DropBox (collaboration) [21, 29, 41].
The U.S. National Institute of Standards and Technology
(NIST) has classified the cloud service models into three 4.2.2 Platform as a Service (PaaS)
categories namely Software as a Service (SaaS), Platform It is an on-demand computing platform and solution
as a Service (PaaS) and Infrastructure as a Service (IaaS). stack [29] service delivery model with a medium level of
Each service model has its own capabilities and corre- abstraction that helps to speed up the development proc-
sponds to serve as an organizational scope and depict a ess. The cloud customers can deploy their own applica-
level of abstraction over the computational environment tions or can personalize applications by using program-
[30, 38, 39]. On the other hand, a research conducted by ming languages and database tools (e.g. J2EE, Java
Youseff, Butrico and Da Silva [2] depicted that the cloud Spring, Ruby on Rails, .NET, PHP) on the development
computing architecture falls into five layers as Cloud Ap- environment of the platform [30, 31]. There is no need to
plications (SaaS), Cloud Software Environment (PaaS), install any developing tools on customer’s computer. The
Cloud Software Infrastructure (Computational Resources PaaS platform provides set of well-defined Application
as Iaas), Storage as (DaaS), Communications as (CaaS), Programming Interfaces (API), with services of configura-
Software Kernel and Hardware or Firmware (HaaS). Each tion management, to interact with other cloud applica-
layer corresponds a level of abstraction that hides all un- tions [39, 40, 42]. While developing application, the de-
derlying complex components and provides simplified velopers take advantages from features as authentication
access of the resources to the cloud customers. Youseff, et services, communications services, automatic scaling,
al., [2] were among the first to attempt a unified ontology load balancing, graphical user interfaces and more [40].
of cloud computing. The novel facilitates a comprehen- The cloud service provider controls and manages all the
sive understanding of cloud technology and its relevant underlying cloud physical infrastructure including hous-
components so to assist the scientific community in con- ing, network, storage, servers and operating systems [29,
structing more efficient functionalities and features for 31], while, the cloud customer controls over the hosting
the cloud (cited by [40]). Followings are the cloud service configurations of the deployed applications and can
models described: avoid concerning about such expenditures [29, 36, 38].
The cloud service provider is responsible to ensure the
4.2.1 Software as a Service (SaaS) security of rented services to the customers and their ap-
It is an on-demand software delivery model with the plications on the platform [39].
highest level of abstraction in the cloud [41]. The custom- Examples of PaaS providers: Microsoft AzureTM
ers’ applications and data are typically stored in the cloud Service platform, Google App Engine, Amazon Web Serv-
and are accessed from various users’ devices (laptop, ices (AWS) Elastic Beanstalk, Amazon SimpleDB or RDS
smart phone, iPad) through thin client interfaces as web for database, SOASTA for development and testing, MS
browser over the network, typically Internet. These Azure or Force.com for general purpose, Amazon SQS,
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

SNS, IBM Cast Iron or Informatica on-demand for inte- width, communication encryption, message delay and
gration and Cloud9 Analytics for business intelligence network monitoring.
[29, 39, 40, 41, 43].
4.2.5 Software Kernel Layer
4.2.3 Infrastructure as a Service (IaaS) Youseff, et al. [2] defined a kernel layer as a basic soft-
It is an on-demand service of an abstracted computing ware management for the physical servers that can be
infrastructure or a virtualization platform [39] that pro- implemented as OS kernel, hypervisor, VM monitor or
vides servers, operating systems, storage and network to cluster middleware.
develop, host and execute applications [29, 30]. Infrastruc-
ture layer, also known as the virtualization layer, is the 4.2.6 Hardware as a Service (HaaS)
most imperative aspect in cloud computing, as the physi- It provides almost similar services as IaaS, only hardware
cal resources are partitioned by virtualization technolo- and VMs are rented in respective services [41]. The HaaS
gies and are used more efficiently, for example, dynamic providers leased out physical resources, which forms the
resource assignment, Xen, KVM and VMware [21]. In- backbone of cloud computing, like computers, servers,
stead of purchasing expensive housing, processing switches, routers, firewalls, load-balancer, power or cool-
power, data-centre, software, complex hardware and ing systems [21]. The cloud users have full control over
network equipments, the cloud customers use the IaaS rented HaaS resources [41] and they perform mainte-
resources as virtualized objects that can be monitored nance include hardware configuration, fault tolerance,
through a service interface. Except basic underlying infra- traffic management, power and cooling resource man-
structure, the cloud customers are responsible for security agement, through the Internet. Typically, HaaS is put into
provisions of the entire setup of applications or the leased operation in a data centre which contains thousands of
infrastructure [39]. They manage and have control over servers in rows and that they are interconnected through
operating systems, storage, deployed applications, host- switches, routers etc. [21].
ing firewalls and possibly some limited network setup Example of HaaS service provider: NewServers [41],
configurations [29, 30, 31]. The cloud users install, config- IBM’s Managed Hosting service [40].
ure and operate the virtual machines rented from the
cloud providers, in which, they can install any operating According to NIST, cloud infrastructure can be divided in
system [41] and their own added services for external or five distinct conceptual layers as shown in Figure 5 be-
internal use [42]. However, the cloud users can also use low:
pre-installed and configured VMs as a base setup by the
IaaS providers [41]. In parallel to computational re-
sources, data storage is also offered as a service, Storage
as a Service, that provides cloud customers on-demand
remote access flexible storage capability. The service pro-
viders typically provide assurances for high availability,
data consistency, redundancy, performance and reliabil-
ity.
Examples are Amazon’s Elastice Block Storage (EBS)
or Simple Storage Service (S3) and Rackspace’s Cloud
Files. In addition, data are also provided as services. For
example Amazon provides human genome or US census
data sets for analytics services [40]. Organizations with
one-time data processing demands and very large data
storage use IaaS services [43]. Others can simply rent a
server time, memory or have an operating system run on
top with their own applications [42]. Fig. 5. Scopes and Control (NIST)
Example of IaaS service providers: Flexiscale, GoGrid,
Amazon Web ServicesTM, Kaavo for cloud broker, Ama-
zon CloudWatch or VPC for services management, Veri- NIST has illustrated the prospects of scope and control
zon CaaS, Rackspace Cloud Servers, Serve Path GoGrid between the cloud providers and customers over each
or Amazon EC2 for compute and Amazon S3 & EBS for layer of service models. Facility layer includes heating, air
storage, Amazon’s Elastic Compute Cloud, Enomalism cooler, ventilation, power, communications and related
Elastic Computing Infrastructure [21, 29, 37, 39, 40, 41, physical plant, while Hardware layer includes computers,
43]. network, storage and related physical computing infra-
structures. Facility and Hardware layers can be both
4.2.4 Communication as a Service (CaaS) termed as HaaS. The Virtualization Infrastructure layer or
Youseff, et al. [2] has also defined a new idea of serv- IaaS includes the logical elements to establish computing
ice as Communication as a Service (CaaS) that will ensure platform, such as, virtual machines, virtual storage and
quality communication services like network security, virtual network. The Platform layer or PaaS includes
virtual overlays for traffic isolation or dedicated band- compilers, libraries, middleware, utilities and other soft-
ware development and deployment tools. The Applica-
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

tion layer or SaaS provides the application for the end- disk, power and more. Business Support provides de-
user or other programs. Cloud computing encompasses ployment, configuration, scheduling, performance re-
with a service provider side and a client side. The left and view, monitoring, software lifecycle management, fault
the right arrows in the Figure 6 represent the range of and logging, user management, authentication, metering
scope and control over the service layers by the cloud and billing [12, 24, 29]. As innovation compels rapid
service provider and customer respectively. In general, standard product development, the cloud stakeholders
the higher the levels of service layers are consumed by will benefit from various ways of interacting service
the customer, the more scope and control the customer models for developing applications and interfaces. Such
has over the system. However, the cloud provider has full flexible capabilities allow cloud customers to customize
control over the two bottom physical layers [30]. the service models that work best for their business re-
Unlike conventional hosting service environments, the quirements.
cloud architecture is very much modular and the layers
are loosely coupled from one another to provide flexibil-
5 CONCLUSION
ity for advancement. Such modularity helps cloud com-
puting to comply with various APIs and thus cutting Cloud based business applications can be an inexpensive
down in management and maintenance expenses [21]. approach as it reduces the overall complexity and saves
Figure 6 shows a Cloud computing architecture in a costs from expensive real-estate, hardware, network
visual form with layers of service models. The figure has bandwidth, cooling equipment, electricity, operational
been visualized from the defined cloud architecture by costs, software licenses, physical security and human re-
different regulatory bodies and research works [2, 3, 12, sources. Cloud service providers bill for each instance,
21, 24, 29, 42, 44]. like pay-as-you-go, so better accountability and transpar-
ency of costs. Unlike traditional computing, it is scalable
and consumes less time to start up. General public can
access these hosted data or application from cloud servers
over Internet through their devices. It delivers quality
services at much lower costs that often discourage small
firms to invest in expensive computing resources. Instead,
they can concentrate more on customer requirements,
which in terms effect in competitive advantage, and pro-
duce solutions that can contribute to make new innova-
tions.

REFERENCES
[1] R. Wang, “Data Centre Best Practices: Managing data with
cloud computing,” InfoWorld, 2013.
[2] L. Youseff, M. Butrico, and D. Da Silva, “Toward a unified on-
tology of cloud computing,” IEEE, pp. 1–10, 2008.
[3] P. Mell and T. Grance, “NIST Definition of Cloud Computing,”
NIST Special Publication 800-145, 2011. [Online]. Available:
http://csrc.nist.gov/publications/nistpubs/800-145/SP800-
145.pdf.
[4] OPC, “Introducing to Cloud Computing,” Office of the Privacy
Commissioner of Canada (OPC), 2011. [Online]. Available:
http://www.priv.gc.ca/resource/fs-fi/02_05_d_51_cc_e.pdf.
[5] L. M. Vaquero, L. Rodero-Merino, J. Caceres, and M. Lindner,
“A break in the clouds: towards a cloud definition,” ACM SIG-
COMM Comput. Commun. Rev., vol. 39, no. 1, pp. 50–55, 2008.
[6] D. Gottfrid, “Self-Service, Prorated Supercomputing Fun!,” The
New York Times, 2007. [Online]. Available:
Fig. 6. Cloud Service Models and Business Support http://open.blogs.nytimes.com/2007/11/01/self-service-
prorated-super-computing-fun/.
[7] J. Varia, “Cloud Architectures,” Amazon Web Services, 2008.
[Online]. Available:
http://jineshvaria.s3.amazonaws.com/public/cloudarchitectur
Terminologies from Figure 6:
es-varia.pdf.
SaaS is built on top of PaaS stacks and provides presenta- [8] K. Swaminathan, “The value of cloud computing to outsourcers
tional platform and self-contained operating environment and their clients,” IBM Software Group, 2009. [Online]. Avail-
through applications and management capabilities. PaaS able:
ftp://ftp.software.ibm.com/software/uk/rational/cloud_com
resides above IaaS stacks and provides programming en- puting_white_paper.pdf.
vironment, system integration and virtualization mid- [9] D. Chappell, “Introducing the Windows Azure platform,”
dleware capabilities. IaaS provides entire infrastructure Microsoft, 2009.
[10] Sun Microsystems, “Introduction to cloud computing architec-
resource stacks and deliver physical and logical connec- ture,” Java.net, 2009. [Online]. Available:
tivity within these resources. Software kernel manages https://java.net/jira/secure/attachment/29265/CloudCompu
software integration over physical hardware. HaaS pro- ting.pdf.
[11] VMware, “The VMware vCloud Solution Enterprise-Class
vides basic physical resources CPUs, routers, memory, Cloud Computing for Any Application, Any Customer,”
 JOURNAL OF COMPUTING AND MANAGEMENT STUDIES VOLUME 1. ISSUE 1. 30TH JANUARY 2017

VMware, 2008. [Online]. Available: [35] Aberdeen Group, “Aberdeen Group,” 2012. [Online]. Available:
http://www.vmware.com/files/pdf/cloud/vmware-cloud- http://www.aberdeen.com/Research/Research- Li-
white-paper.pdf. brary.aspx?search=private could.
[12] Z. Xiao and Y. Xiao, “Security and Privacy in Cloud Comput- [36] H. Susanto, M. N. Almunawar, and C. C. Kang, “A Review of
ing,” IEEE Commun. Surv. Tutorials, vol. 15, no. 2, pp. 843–859, Cloud Computing Evolution Individual and Business Perspec-
Jan. 2013. tive,” Academia, 2012.
[13] J. Voas and J. Zhang, “Cloud Computing: New Wine or Just a [37] T. H. Noor, Q. Z. Sheng, S. Zeadally, and J. Yu, “Trust Man-
New Bottle?,” IEEE Comput. Soc., vol. 11, no. 2, pp. 15–17, Mar. agement of Services in Cloud Environments  : Obstacles and So-
2009. lutions,” vol. 46, no. 1, 2013.
[14] L. Mei, W. K. Chan, and T. H. Tse, “A Tale of Clouds: Paradigm [38] E. Aguiar, Y. Zhang, and M. Blanton, “An Overview of Issues
Comparisons and Some Thoughts on Research Issues,” in IEEE and Recent Developments in Cloud Computing and Storage
Asia-Pacific Services Computing Conference (APSCC 2008), 2008. Security,” in High Performance Cloud Auditing and Applications,
[15] Y. Y. Alshamaileh, “An empirical investigation of factors affect- Y. Choi, K. Han, and S. Song, Eds. Springer, 2014.
ing cloud computing adoption among SMEs in North Eng- [39] N. R. Putri and M. C. Mganga, “Enhancing Information Se-
land,” Newcastle University of Business School, 2013. curity in Cloud Computing Services using SLA Based Metrics,”
[16] K. Hwang, “Massively Distributed Systems  : From Grids and Blekinge Institute of Technology, 2011.
P2P to Clouds,” in Advances in Grid and Pervasive Computing, [40] M. Böhm, S. Leimeister, C. Riedl, and H. Krcmar, “Cloud Com-
vol. 5036, Berlin, Heidelberg: Springer, 2008, p. p1. puting - Outsourcing 2.0,” 2009.
[17] G. Motta, N. Sfondrini, and N. Sacco, “Cloud Computing: An [41] P. Egwutuoha, D. Schragl, and R. Calvo, “A Brief Review of
architectural and technological overview,” IEEE Xplore, 2012. Cloud Computing, Challenges and Potential Solutions,” Parallel
[18] Foster, Y. Zhao, I. Raicu, and S. Lu, “Cloud computing and grid Cloud Comput., vol. 2, no. 1, Jan. 2013.
computing 360-degree compared,” IEEE, 2008. [Online]. Avail- [42] S. Mewada, U. Singh, and P. Sharma, “Security Enhancement in
able: Cloud Computing,” Int. J. Sci. Res. Comput. Sci. Eng., vol. 1, no.
https://scholar.google.com/citations?view_op=view_citation& 1, 2013.
hl=en&user=jE73HYAAAAAJ&citation_for_view=jE73HYAAA [43] W. L. Hsu, “Conceptual Framework of Cloud Computing Gov-
AAJ:u5HHmVD_uO8C. ernance Model – An Education Perspective,” IEEE Technol. Eng.
[19] R. Buyya and S. Venugopal, “A Gentle Introduction to Grid Educ., vol. 7, no. 2, pp. 12–16, 2012.
computing and technology,” Computer Soceity of India, 2005. [44] X. Chen, G. B. Wills, L. Gilbert, and D. Bacigalupo, “Using
[Online]. Available: Cloud for research: A Technical Review,” TechiRes, 2010. [On-
http://www.cloudbus.org/papers/GridIntro-CSI2005.pdf line]. Available:
[20] K. Hashizume, “A reference architecture for cloud computing http://tecires.ecs.soton.ac.uk/docs/TeciRes_Technical_Report.
and its security applications,” Florida Atlantic University, 2013. pdf
[21] Q. Zhang, L. Cheng, and R. Boutaba, “Cloud computing: state-
of-the-art and research challenges,” Brazilian Comput. Soc., pp.
7–18, 2010.
[22] S. Gadia, “Cloud Computing  : An Auditor’s Perspective,” Ishrat Ahmad, PhD candidate at London School of Commerce/Cardiff
ISACA J., vol. 6, pp. 1–5, 2009. Metropolitan University, UK. Email: [email protected]
[23] E. Schmidt, “Conversation with Eric Schmidt hosted by Danny
Sullivan,” Search Engine Strategies Conference, 2006. [Online]. Dr. Humayun Bakht, Director of Studies at London School of Com-
Available: merce/Cardiff Metropolitan University and Principal Examiner Professional
http://www.google.com/press/podium/ses2006.html Issues in IT (PITT) at NCC Education. Email: [email protected]
[24] L. Qian, Z. Luo, Y. Du, and L. Guo, “Cloud Computing  : An
Overview,” pp. 626–631, 2009. Dr. Uma Mohan, Supervisor and Senior Programme Leader at London
[25] Z. Enslin, “Cloud computing  : COBIT-mapped benefits, risks School of Commerce, UK. Email: [email protected]
and controls for consumer enterprises,” Stellenbosch  : Stellen-
bosch University, 2012.
[26] S. Marston, Z. Li, S. Bandyopadhyay, J. Zhang, and A. Ghalsasi,
“Cloud Computing - The business perspective,” Decision sup-
port systems, 2011. [Online]. Available:
http://www.acc.ncku.edu.tw/chinese/faculty/shulc/courses/
IT_Competitive_Advantage/assignments/2012-fall/cloud-
computing-business-perspective.pdf
[27] J. Hiner, “Apple cloud vs. Google cloud: The philosophical
differences - TechRepublic,” Techrepublic, 2011. [Online]. Avail-
able: http://www.techrepublic.com/blog/tech-sanity-
check/apple-cloud-vs-google-cloud-the-philosophical-
differences/?tag=nl.e098
[28] B. Hayes, “Cloud Computing,” Communications of the ACM, vol.
51, no. 7, 2008.
[29] J. Archer, D. Cullinane, N. Puhlman, A. Boehme, P. Kurtz, and
J. Reavis, “Cloud Security Alliance Top Threats V1.0,” 2010.
[30] W. Jansen and T. Grance, “NIST Guidelines on Security and
Privacy in Public Cloud Computing,” 2011.
[31] D. Zissis and D. Lekkas, “Addressing cloud computing security
issues,” Futur. Gener. Comput. Syst., vol. 28, no. 3, pp. 583–592,
Mar. 2012.
[32] Omotunde, O. Awodele, O. Kuyoro, and C. Ajaegbu, “Survey
of Cloud Computing Issues at Implementation Level,” CIS J.,
vol. 4, pp. 91–96, 2013.
[33] M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. Katz, A.
Konwinski, G. Lee, D. Patterson, A. Rabkin, and I. Stoica, “A
View of Cloud Computing.,” Communications of the ACM, vol.
53, no. 4, pp. 50–58, 2010.
[34] T. Dillon, E. Chang, and C. Wu, “Cloud computing: Issues and
challenges,” in 24th IEEE International Conference on Advanced In-
formation Networking and Applications (AINA ’10), 2010, pp. 27–
33.

View publication stats