Cloud Computing: What Everyone Should Know

It would be hard to avoid the all the talk that surrounds cloud computing. Gartner recently put it at the top of its "Hype Cycle", near the “peak of inflated expectations” and quickly approaching the dreaded “trough of disillusionment”. Even mainstream publications like the New York Times and The Economist have written lengthy articles on it. Funny, then, that if you were to ask two people to define cloud computing you’ll likely end up with two different definitions. Sometimes, they’ll be wildly different.

This is as much a case of the perception that we’re dealing with something entirely new, as it is a phenomenon of the 24/7 news media cycle combined with user-generated content and commentary. Analysts, industry pundits, writers and bloggers all have to find a way to rise above the cacophony of voices expounding on all things technology. Often, that results in putting down one person’s commentary over his or her own.

As you’ll see, cloud computing is not something that was created out of ether or deep in someone’s lab. At most, it’s a new term for an evolving paradigm.

What is often absent in the discussion of cloud computing is context. That makes it easy to pick apart as Silicon Valley’s latest trend. But, that would be the wrong assumption.

Now, having spent a lot of time reading articles (and watching videos) of some very bright people argue about what cloud computing is and isn’t, I’m here to cut to the chase. I’m going to call on two resources to make my case and let you decide whether I’ve proven my case and distilled cloud computing down to its essential parts.

Simon Wardley, a strategist at Canonical, perhaps more so than anyone else I have come across, has eloquently and succinctly explained cloud computing from an historical perspective. His recent keynote at OSCON ’09, one of the year’s biggest technology conferences, was masterful in its ability to take the heat out of such a charged topic and place it within the appropriate context. That’s all the more important because with context comes rationality and with it a pragmatic approach to dealing with forces of change.

Wardley makes it clear that cloud computing isn’t about technology alone, but rather a consequence and evolution of several factors working together. Similar to the Industrial Revolution, it is being driven by cultural and technological factors. Take the concept of utility computing and the changing attitude that sees IT’s value more in terms of a cost of doing business, rather than one of strategic value (in its current state).

Throw in the constant pressure of product commoditization and newer technology, like virtualization (a building block for cloud computing), and you have the makings of a honest-to-goodness transformation.

It’s not a trend or the latest fashion. It is, as he states, a highly disruptive shift of IT from one that was product-based to one that is services-based – one whose time has come. He admonishes organizations to change and adapt as the balance between innovation and commoditization forces competitive pressure on organizations of all sizes.

Cloud computing is a generic term used to describe the disruptive transformation in IT towards a services-based economy, driven by a set of economic, cultural and technological conditions.

- Simon Wardley, Canonical

So, that gives us a baseline and puts things in context. But, to flesh out the concept in more detail, I’d like to call on another excellent resource: The National Institute of Standards and Technology (NIST). The NIST is an agency within the Department of Commerce whose “role in cloud computing is to promote the effective and secure use of the technology within government and industry by providing technical guidance and promoting standards”. These are your tax dollars at work. And they just happen to have produced, through countless revisions, one of the more complete definitions available.

Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.

- NIST Definition of Cloud Computing

You can read their definition by following the link provided above. Suffice it to say, the NIST has done a great job and service to the industry by explaining what is (and by definition, what is not) cloud computing.

Follow the essential characteristics and you’ll quickly be able to tell those who are truly delivering cloud computing services from those who have merely slapped a cool marketing term on their products. In short, those characteristics include: On-demand self-service; Broad network access; Resource pooling; Rapid elasticity; and Measured service.

Another important element is the ability to distinguish between service models, or types of providers, in order to distinguish and segment the cloud provider landscape. Accordingly, the cloud provider landscape can be grouped into three categories:

• Software as a Service (SaaS) is delivered on a pay-per-use basis to provide end-users with a specific business application, like CRM, accounting, e-mail, or collaboration. Examples include Salesforce.com, Google Apps, and NetSuite.
• Platform-as-a-Service (PaaS) delivers an open platform for developers to build and scale their applications. Force.com from Salesforce and Google App Engine are examples of two cloud platform providers.
• Infrastructure-as-a-Service (IaaS) provides a complete set of resources on which developers can run their applications. Examples include Amazon EC2, 3tera and Rackspace.

Then, there are different ways that cloud computing can be deployed. When people think of cloud computing they often think of Amazon and its public cloud offerings like Amazon EC2 (for on-demand servers) or Amazon S3 (for on-demand storage). At the other end of the spectrum are private clouds, those where the cloud is the sole property of the company for which it was deployed. Often built on-premise and with existing data center assets, they are designed to marry the benefits of cloud computing with a company’s policy control over data security and privacy. In between, so to speak, are hybrid clouds that aim to combine public and private clouds that “are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting for load-balancing between clouds)”.

So, there you have it: cloud computing in both the broad bush strokes and with added detail for clarity and consistency. As a result, cloud computing should be seen in the proper context – as part of an evolving and transformational paradigm – and into component parts that make it easier to understand and identify.