What is the Cloud?
There are a lot of “cloud” companies out there, each with their own definition of what the cloud is and what it means to you. I am proud to say that I worked for a company that had one of the first true cloud products. The company was Concentric Networks (now XO Communications) and the product was Clustered Hosting. They did not quite meet the definition that I am about to give you, but even today, they are closer to the concept than many companies who, today, call themselves “cloud based”.
First let me give you a very high level concept of an application supported by its underlying infrastructure. All of the lower levels support the application that lives at the top.
- The Application – The Application is the cloud storage application that you use, or your website, or your CRM (like Insightly).
- OS and Management – This refers to the duties that are provided by the IT department – updates, backups, remediation, etc.
- Server and Storage – These are the sum total of the computing devices that deliver the application.
- Network – This is the network delivery to the upper layers that deliver the Application.
- Data Center – This is the physical location that delivers the Application.
In this model, the failure of a lower level affects everything above it. If the network goes down, then everything above it is in effective failure. What good are servers if the network is down. You get the idea.
Lets first define a pure definition of cloud computing and then compare it to some application delivery scenarios that are defined as “cloud” but truly are not.
A true cloud application is one where the user has no responsibility to manage any of the application layers.
Using Insightly as the example, of course the user would never have any responsibility for any layer because they did not and can not affect any layer. They just don’t have access, so they don’t have responsibility. That means that the user is truly free of any other expense or obligation to manage the application. A true cost savings.
A true could application has maximum resiliency to disasters, outages, regional problems, DDOS attacks, security issues, etc.
A cloud application is delivered by a cluster of CPU’s – not just one. When one of those CPU’s fails, the others compensate. For problems like these there is no diminishment of computing power for the end user. User data is stored and backed up on a highly redundant Storage Area Network (SAN). The cloud means that the user does not have to worry about losing data. There are multiple network paths in and out of the data center(s), so there is no single point of failure for the user to get to the application. No backhoe risk here. Having multiple data center redundancy is much more complex and much more expensive. True cloud providers replicate data across data centers. The failure of one data center means another delivers the application. Not easy to do.
A cloud delivered application has infinite resources.
This may sound unreasonable, but to a single user, it should be a reality. If you manage the cloud computing platform that has multiple data centers and a massive, redundant infrastructure, their doubling or tripling of traffic will be a blip on your usage meter. Your vision into usage patterns over hundreds or thousands of users gives you the ability to anticipate and incrementally handle increased usage.
Knowing this, let’s examine some “cloud” providers that really are not.