My eighth-grade English teacher, Mrs. Pedrick, probably would have a field day with her red pen if she ever read many of the IT-related blogs out there – including mine. One of her pet peeves was the loading up of sentences with redundancies; most thoughts could be expressed neatly and cleanly with an economy of words. Consider phrases such as “free gift,” “all-time record,” “foreign imports,” “advance planning,” “join together” and “new recruit.”
Perhaps “cloud computing” has become one of those redundant phrases as well. I recently had the opportunity to moderate a panel discussion on cloud at IBM's Impact conference. One of the panelists, Dr. Angel Diaz, vice president of software standards & cloud for IBM, predicted that we may not even be using the term “cloud” within the next five years – any and all forms of computing will be taking advantage of a mix of network and local resources.
In the meantime, another IBMer, Chris Dotson, wants to make the point that cloud, as it exists today – occurs in many different forms. He provides four reasons why the term “cloud” gets brutally misused across the industry:
- There isn’t a single cloud: Every vendor has a different type of offerings, such as Apple's iCloud or Amazon's EC2 Cloud or IBM’s SmartCloud. “Even within an organization, it is certainly possible that there would be different private clouds for different purposes,” he adds.
- There are many types of cloud: Consider all the different purposes and architectures cloud performs:“private development/test compute cloud, a private storage cloud, a private desktop cloud, analytic cloud, and many others.
- Many people think that “the cloud” is magic: Clouds just don't automatically run every workload you throw at them; it still takes a lot of integration and development work. “Cloud computing is simply a different model with the same old mundane computers beneath it,” says Dotson.
- Talking about “the cloud” implies that it’s a thing rather than a service model: “To me, the most important piece is on demand self-service, meaning that you can use services in an automated fashion without waiting for another human to help you unless something goes wrong. Practically anything that meets these requirements can be used or sold in a cloud model, even some things that have been around for years! What’s different here is that the number of these services is growing so quickly, the interfaces between them are standardizing somewhat so that different services can be swapped in and out for different purposes, and that these services are able to make use of other cloud services in an automated fashion. The biggest impact of cloud computing might not be in humans requesting services, but in cloud agents requesting services on behalf of humans!”