Why is "The Cloud" still unclear?

After more than five years of using the term "The Cloud", technology vendors, resellers, service providers, analysts, and IT professionals still use inconsistent definitions for cloud computing (i.e. "The Cloud").  There are plenty of reasons for this lack of consistency, but the end result is confusion for many organizations seeking the benefits of cloud computing. The intent of this post is to break down cloud computing into manageable segments to provide leaders of organizations (business, education, government, etc.) a foundation to converse with IT staff, consultants and vendors on potential cloud solutions.

Establishing a Cloud Foundation

To explain cloud computing I'll use simplified definitions based on a well-established set of terminology published by the National Institute of Standards & Technology (NIST). NIST offers multiple cloud publications on their site including a condensed paper, NIST Definition of Cloud Computing.

NIST breaks down cloud computing into three parts:

• Service Models: Three different models provide computing hardware and software resources as a service to consumers (consumers are customers/end users). When conceptualizing computing resources delivered as a service, it often helps to think in terms of a utility such as water or electricity.
• Essential Characteristics: Five characteristics that differentiate cloud computing from the service models. When all five characteristics are combined with one of the three service models a cloud infrastructure is established.
• Deployment Models: Once a cloud infrastructure is established it must be deployed (delivered) to the consumer. There are four models for deploying cloud infrastructure, and those that deliver a cloud infrastructure to a consumer are known as providers.

As each part of the cloud is explained, keep this equation in mind.

 

The Service Models

The three service models are differentiated by the level of hardware and software resources delivered. The first service model delivers the most basic level of resources, which is the hardware. Each of the subsequent models offer additional resources on top of what was offered in the previous model.

• Infrastructure as a Service (IaaS): Delivers hardware resources such as servers.
• Platform as a Service (PaaS): Delivers the server hardware, plus the operating system that runs on the server, such as Microsoft's Windows Server.
• Software as a Service (SaaS): Delivers the server hardware, the operating system that runs on the server, plus the application that runs on the operating system, such as Microsoft Dynamics GP.

These service models are the foundation of cloud computing, but were established practices prior to the inception of the cloud.

The Essential Characteristics

The cloud infrastructure is established by combining one of the above service models with all five cloud characteristics.

• On-demand self-service: The ability for a consumer (end user) to provision computing resources automatically without human intervention, typically done through a web portal.
• Broad network access: The ability for a consumer to access computing resources (often applications or documents) from any device (laptop, tablet, smart phone, etc.) from any location.
• Resource pooling: The provider’s computing resources (storage, servers, network bandwidth, etc.) are pooled together and dynamically assigned and reassigned to multiple consumers based on demand.
• Rapid elasticity: The ability to quickly scale up or scale down the pool of computing resources to meet consumer demand.
• Measured service: The pooled resources are metered to capture consumer usage, and automatically controlled and optimized based on usage.

These characteristics are what define cloud infrastructure from just a service model offering.

The Deployment Models

Once the cloud infrastructure is established by combining the five characteristics with a service model, it is must be deployed to consumers using one of four models.

• Public Cloud: Cloud infrastructure is built and offered by a provider and shared by multiple seperate organizations. The cloud infrastructure exists on premise (data centers) of the provider. Examples of public cloud providers include Amazon, Google and Microsoft.
• Private Cloud: Cloud infrastructure is built and offered exclusively for one organization. It may be owned, managed, and operated by the organization, a third party, or some combination of them, and it may exist on or off premise. One example is an IT department builds their own on premise cloud using a 3rd party platform such as VMware. A second example is a provider such as Amazon builds a dedicated cloud for an organization on Amazon's platform hosted off premise in Amazon's data centers.
• Community Cloud: Similar to a public cloud, but the cloud infrastructure is built and shared by a community of organizations with shared concerns. A common example is a health care community cloud where providers have the shared goal of complying with HIPPA standards.
• Hybrid Cloud: A combination of any two or even all three of the above models. The most common example is an organization builds a private cloud on their own premise, but during high traffic times the added workload is transferred out to a public cloud provider.

When the cloud infrastructure is deployed using one of the above models, cloud computing is achieved.

Summary

The most important thing for any organization evaluating a cloud offering is to keep it simple. Request the product reseller or service provider map their offering to each part of the cloud. Keeping it simple helps to ensure the product reseller or service provider is not overselling by slapping "cloud" onto their solution. Too often sales people (intentionally or unintentionally) use the term cloud to describe their product or service, when it doesn't have all of the characteristics of cloud computing. It also helps to ensure the consumer doesn't buy more than they need. Sometimes a customer doesn't require all of the characteristics of the cloud, and a simpler, less costly solution will meet their business requirements.

I can't answer why after more than five years the terminology to define the cloud is still inconsistent. Maybe it is due to an established body such as NIST only publishing their cloud computing terminology within the last 18 months. Or it could be due to vendors and service providers adopting only the aspects of the cloud that fit their product or service. One thing is certain, there is no single answer. I'd like to hear your thoughts on the whys, or general comments on this post. Please stay tuned for future posts that continue the cloud conversation, including my next post which challenges the need for cloud computing.