Eventually, your application needs to run on a machine somewhere in the world and not just on your local development machine. For many years, when someone wanted to run a web application and make it available to the external world, they would buy designated hardware and spent time maintaining it. Over the years, the hosting model has changed and moved to external hosting that, besides running the machine at some data center outside, felt very similar to what you had when you ran it locally—however, the big part was that you didn't have to worry about the machine maintenance, the server room air conditioning, and the procedure that should be performed in case of hardware failure. Then came the cloud. 

The National Institute For Standards and Technology (NIST) defines the cloud like this:

NIST defines the following five characteristics as essential for making the cloud what it is:

• On-demand self-service: The ability to provision and remove resources with minimal effort
• Broad network access: The ability to manage your resources from any location and the ability to work with resources in a global scale
• Resource pooling: Virtualization of resources that allow you to consider the available capacity as infinite
• Rapid elasticity: The ability to add or remove resources based on the real usage without prior investments
• Measured service: Being able to retrieve metrics on the resources you use to make decisions based on real data

One of the things that hides in these characteristics is that the cloud allows you to automate many (if not all) the operations needed to run your application, and this is made possible above all because of the virtualization that the cloud hypervisor uses.

In the rest of this book, I will delve deeper into how those characteristics come in to play, and in this chapter, I will explain the service models and the connection between the cloud, virtualization, and automation.