Cloud Computing - Cloud Data Management - page 14

the three service models from the perspective of the providers' view. The following
definitions of the three models combines the two perspectives [ 181 , 225 ], in the hope
of showing the whole picture.

1. Infrastructure as a Service : Through virtualization, the provider is capable of
splitting, assigning, and dynamically resizing the cloud resources including
processing, storage, networks, and other fundamental computing resources to
build virtualized systems as requested by customers. Therefore, the customer
is able to deploy and run arbitrary operating systems and applications. The
customer does not need to deploy the underlying cloud infrastructure but has
control over which operating systems, storage options, and deployed applications
to deploy with possibly limited control of select networking components. The
typical providers are Amazon Elastic Compute Cloud (EC2) [ 4 ] and GoGrid [ 17 ].
2. Platform as a Service : The provider offers an additional abstraction level, which
is a software platform on which the system runs. The change of the cloud
resources including network, servers, operating systems, or storage is made in
a transparent manner. The customer does not need to deploy the cloud resources,
but has control over the deployed applications and possibly application hosting
environment configurations. Three platforms are well-known in this domain,
namely Google App Engine [ 19 ], Microsoft Windows Azure Platform [ 37 ], and
Heroku [ 28 ] which is a platform built on top of Amazon EC2. The first one
offers Python, Java, and Go as programming platforms. The second one supports
languages in .NET Framework, Java, PHP, Python, and Node.js. While the third
one is compatible with Ruby, Node.js, Clojure, Java, Python, and Scala.
3. Software as a Service : The provider provides services of potential interest to a
wide variety of customers hosted in its cloud infrastructure. The services are
accessible from various client devices through a thin client interface such as a
web browser. The customer does not need to manage the cloud resources or
even individual application capabilities. The customer could, possibly, be granted
limited user-specific application configuration settings. A variety of services,
operating as Software as a Service, are available in the Internet, including
Salesforce.com [ 43 ], Google Apps [ 21 ], and Zoho [ 55 ].

2.4 Cloud Deployment Models

The guideline also defines four types of cloud deployment models [ 181 ], which are
described as follows:

1. Private cloud . A cloud that is used exclusively by one organization. It may be
managed by the organization or a third party and may exist on premise or off
premise. A private cloud offers the highest degree of control over performance,
reliability and security. However, they are often criticized for being similar
to traditional proprietary server farms and do not provide benefits such as no
up-front capital costs.