Testing and Deployment: Quality Is King - Deploying and Managing a Cloud Infrastructure - page 288

This chapter covers the deployment and testing of cloud com-
puting models. The most popular are private, public, and
hybrid models. These models present options for deploying
applications in the cloud. Each model presents a different set of options, such as deployed
services, security, and accessibility, among others. Moreover, each model also presents dif-
ferent suitability criteria, including location of deployment (on site or off site), sensitivity
of data (confidentiality and data security), and costs. This chapter covers cloud computing
models and strategies to extend virtualization outside the data center and to the cloud.
Overview of Deployment Models
With emerging cloud implementations, organizations look to expand their businesses and
scope, and they need to take advantage of local, offsite, and public cloud offerings. The exist-
ing cloud models derive their types from methods of deployment as well as client usage.
National Institute of Standards and Technology (NIST) Special Publication (SP) 800-145
describes four basic models for cloud deployment, as described in the following sections.
Private Cloud
Private clouds reside on a private network that is owned, managed, and operated by the
organization itself. Private clouds provide internal users with immediate access to comput-
ing resources. Provisioning and scaling a collection of resources is usually performed using
a web interface. The cloud computing resources can be situated on or off site and can be
managed by an in-house team or a third party.
One of the biggest advantages of having a private cloud is security. In fact, this model
addresses the security and privacy issues, due to shared resources and multitenancy, inher-
ent in other cloud models. Because a private cloud is behind the firewall, it is subject to the
physical, electronic, and procedural security measures of the organization. Moreover, it
offers a higher degree of security for intellectual property (IP) such as code and data.
A private cloud gives network administrators the ability to control performance of the
physical machines, thus offering an opportunity to improve efficiency and reduce costs.
However, the biggest downside of this model is heavy setup costs and recurring operational
and maintenance costs. In addition, the setup requires special skilled resources such as cloud
architect and IT engineers.