Cloud as a Computer - Advanced Design Approaches to Emerging Software Systems - page 244

that did not exist in the olden days. A cloud can
potentially include virtually everything that we can
access over the Internet - hardware and software
included. That is almost synonymous with infinite
computing. Include in the equation, the prolifera-
tion of the inexpensive mobile devices that can
access the Internet from virtually anywhere in the
world, and we have a ubiquitous, infinite capacity
cloud as a computer. Some authors (Durkee 2010)
consider cloud computing as timesharing reborn.
It is indeed true that cloud computing has similar
impact as timesharing in the 1980s.

and executed on these images. Clients can be rest
assured that their capacity needs are automatically
met, without budgeting for maximum capacity, as
is typically done with datacenters. They can still
pay for just what they used.
The greatest advantage of Cloud Computing is
that anyone who can afford for operating expenses
can jump start into a business by deploying their
applications in the cloud, without having to invest
in the capital intensive IT equipment. The Cloud
Computing business model changes Capital Ex-
pense (CapEx) into Operational Expense (OpEx).
Imagine businesses having to setup their own
power generators to operate, as they used to do
in early 1900s. Cloud Computing does the same
thing to IT as the Power Utility companies did
to electricity.

Business Model

Cloud Computing is based on Cost Associativity
(Michael et al. 2009), which is the concept of pay-
ing the same price for 'N' units of computational
power for 1 hour or 1 unit of computational power
for 'N' hours. This allows the supply and demand
of computing resources to be elastic. Economists
will readily recognize what elasticity can do in
the market. Price elasticity is a very important
economic parameter of a product. The elasticity in
Cloud Computing refers to the fact that users can
vary their demand of the computing power and the
cloud providers oblige their demand accordingly.
Statistical multiplexing is used to handle the vary-
ing demands from clients, increasing throughput.
This commercially available elasticity of
resources, without paying a huge premium is
completely new to IT. Like other virtual computing
schemes such as virtual memory, the virtualiza-
tion in the cloud creates an illusion of unlimited
computing power to the clients connecting to it.
Virtualization can take the form of hardware or
software virtualization. Hardware virtualization
provides transparency with respect to the backend
configuration of machine resources and enables
plug-and-play mode of hardware resources. Soft-
ware virtualization, on the other hand, refers to
provisioning of software images of the needed
software configuration in terms of Operating
Systems and applications. Code can be assembled

Cloud Ecosystem and Tools

Cloud Computing, as a technology is not com-
pletely new. It can be viewed as a descendant of
a host of other technologies like Client-Server
Computing, Software As A Service (SaaS), and
Virtualization. Though the technologies existed,
Cloud Computing is unique in bringing them all to-
gether. There are a number of XaaS (X as a Service)
terms in the cloud computing paradigm today, but
all stake holders agree that Cloud Computing can
be broadly divided into 3 layers or service models:
SaaS (Software as a Service), PaaS (Platform as
a Service), and IaaS (Infrastructure as a Service).
IaaS is the layer Saas provides ready-made ap-
plications to the end user that are developed and
run on PaaS or can use IaaS directly. There are
several other XaaS, as listed in the introduction,
but we shall eschew them for better clarity, as the
authors of (Michael et al. 2009) did.
The whole idea of XaaS is to harness the various
resources available in the cloud. These resources
can be raw hardware such as processors and stor-
age, software such as development tools, applica-
tion servers, and databases. Other resources can
be applications such as word processors and other