Cloud Computing: History, Service Models, Architecture, Security, and Limitations

Cloud computing is an IT paradigm that enables ubiquitous means of entry to communal pools of consignable system resources and services of high levels which can be provisioned rapidly with minimal supervisory efforts, normally over the internet. This presentation covers the history of cloud computing, service models (IaaS, PaaS, SaaS), architecture, security, emerging trends, and limitations.

CLOUD COMPUTING
Cloud computing refers to an IT (Information
Technology) paradigm which enables ubiquitous
means of entry to communal pools of consignable
system resources and services of high levels which
can be provisioned rapidly with minimal
supervisory efforts, normally over the internet. It
relies on sharing of resources to accomplish
economies of scale and coherence, just like in
public utility.

History of Cloud Computing
Cloud computing began in the early 1960’s with
introduction of RJE (Remote Job Entry) associated
with the then large vendors such as DEC and
IBM. Full-time-sharing solutions were already in
the industry by mid-1970’s on platforms like
Cambridge CTSS and Multics (on GE hardware).
Telecommunication companies fanned into this
industry in the 1990’s.

Telecommunication companies that offered
predominantly dedicated point-to-point
circuits of data, started offering VPN (Virtual
Private Network) services with tantamount
quality of service, though at lower costs.
Through switching traffic as deemed fit in
balancing server use, they could put to use
most network bandwidth most effectively.They started using cloud symbol denoting demarcation point between what
consumers were responsible for, and what the provider was susceptive to.

SERVICE MODELS
On the left is a breakdown of cloud computing
service models systematized as layers forming
a stack. Providers of cloud-computing offer
their services depending on the models, which
are three according to international standards
(NIST): PaaS (Platform as a Service), SaaS
(Software as a Service) and IaaS
(Infrastructure as a service).

Infrastructure as a Service
(SaaS)
IaaS is a service that provides APIs for
dereferencing of various details (low-level)
of elemental network infrastructure online
e.g., location, physical computing
resources, scaling, partitioning, backup,
security etc. Hypervisors like Oracle Virtual
Box, Xen, Oracle Vm run virtual machines as
guests. Such pools of hypervisors in a cloud
system support the ability to scale services

The cloud operational system also supports
pragmatic virtual machines. A linux
container will run in an isolated partition of
a Linux kernel which is running directly on
physical hardware. Cgroups and
namespaces, linux kernel technologies are
used in isolation, securing and management
of containers. Containerization is better
than virtualization in terms of performance

Platform as a Service
Vendors of PaaS offer an appropriate environment,
programming language execution environment, for
application development to developers. A provider
develops standards and the toolkit for development
and the channels for payment and distribution.
Developers can therefore develop, test and run
their applications on a cloud based platform
without having to cover costs of buying physical
hardware needed to run their applications

Several data and integration
management providers have
adopted PaaS applications as
models of delivery for data
solutions e.g., Data Platform as a
Service (dPaaS) and Integration
Platform as a Service (iPaaS).
dPaaS delivers data management
products as fully managed
services. iPaaS on the other hand
facilitates development and

Software as a Service (SaaS)
SaaS also called an ‘on-demand software’ is a
model for software delivery and licensing and
works on a subscription basis, licensing of
software is subscribed for, and all this is
hosted centrally. Providers of the cloud
platform install and operate the software on
cloud, and users are allowed access through
cloud clients. Users don’t manage the platform
or infrastructure on which the application runs.
They therefore don’t have to install the

Tasks are cloned onto various virtual
machines to meet the changing
demand on work. Load balancers are
responsible for distributing the work
over several virtual machines. There is
transparency in the process for the
user who only sees one access point.
The process followed when bidding for
space and access to the cloud platform
is shown in the picture.

Deployment Models
A public cloud is where services are offered
over an open network for public use.
A private cloud is that which is developed
for a sole organization regardless of
whether it is hosted externally or internally,
or managed by a third party or internally.
A hybrid cloud is composed of two or more
clouds which maintain their entity identity
but are still bound together.

Cloud Computing Architecture
This refers to components or sub-
components needed for cloud
computing and involves various
cloud components in
communication over a coupling
mechanism that is lose e.g., a
messaging queue. They include
front-end platforms like Cloud
Client Platform (fat client, thin
client, mobile device), Cloud
storage (database), Cloud

Security and Privacy
A service provider has authority to access the
data at any time. This information can be
shared with third parties like law enforcers
without warrants. This is normally indicated in
the privacy policy which users agree to when
they sign up for the cloud computing services.
However, users can also encrypt the data they
store on the cloud to restrict access to
unauthorized parties.

Emerging Trends in Cloud
Computing
Software paradigm has been revolutionized by cloud
computing as it has been changed into a paradigm that
is service oriented where cloud resources and software
are a service. Microsoft, Amazon and other tech
companies are investing billions of their returns in
cloud computing.
Organizations are resorting to collecting and storing
business intelligence in a secure location (cloud).

Limitations of Cloud
Computing
• Security and privacy: Any technology that is internet based is susceptive
to hacking by other parties. Code Space for example, has its AWS EC2
console hacked and data deleted forcing them to close down.
• Limited flexibility and control: Users have no control on the hosting
infrastructure. They have limited control over their developed
applications.
• Platform dependencies.
• Downtime.