Testing and Deployment: Quality Is King - Deploying and Managing a Cloud Infrastructure - page 322

The chapter included cloud architectures and their technical advantages in terms of
resource management, resource allocation, task coordination, job scheduling, and scalabil-
ity. It also included business benefits such as pay per use and capital cost savings.
The chapter then discussed cloud deployment options while keeping in perspective impor-
tant aspects such as service provider API and compatibility with applications. We further
discussed and compared popular commercial offerings, including Windows Azure, Amazon
EC2, and Rackspace. We also included discussions of deployment steps and testing and
monitoring deployed applications.
The chapter then walked through the steps of creating and deploying applications on
Windows Azure and Amazon EC2 cloud services. Last, we discussed organizational best
practices for a coherent cloud deployment model to ensure security and quality.
Chapter Essentials
Cloud Models and Management Strategies Cloud models are derived from their deployment
methodologies and their use cases. They help standardize specifications for organizations to
adopt and migrate for cloud computing. The cloud management strategies help to monitor
tasks related to performance, security, and disaster recovery.
Cloud Architectures It is the system architecture of software systems that are involved
in the delivery of cloud services. Cloud architectures use the underlying infrastructure to
perform a specific job by drawing the resources, relinquishing the unneeded resources, and
often disposing them once the job is complete. Cloud architectures solve most of the tra-
ditional data processing problems such as commissioning machines, provisioning applica-
tions, distributing and coordinating tasks, and ensuring recovery mechanisms.
Deployment Testing and Monitoring Testing the software during the development process
of a cloud application using local compute instances and storage helps the development and
testing teams avoid most of the issues before the staging process even begins. Running unit
and ad hoc tests ensures that the components perform as they are intended.