Cloud Is About Technology

I continue to hear the question "what is cloud" being asked by people who are pretty well versed in IT. I am talking about people who are working in various levels of IT today and have been for years. I also hear many professionals (vendors, analysts, press, experts, bloggers, etc) who have been involved and following the cloud cycle lamenting that every article, presentation, paper and blog has to define cloud, that all the definitions of cloud are different, and these professionals are tired of the definitions. "We've done this already. Let's move past it!" they collectively cry and either point to their own cloud definition or to NIST's.  Ironic or what? The answers to what cloud is and how it is implemented is every bit important and relevant today as it was three years ago.

Scott Stewart wrote a short blog, "There is no real definition for Cloud Computing," where he says many of the things that I have thought for a long time about how and why the definition of cloud computing is cloudy at best. That blog is an extract of a longer thought piece he is writing, but he wraps up with this: "Sadly they may be missing the real point that this cloud computing disruption is not actually about a technology, it is not about a product, it is not a service offering, it is not something that we have always done, it is not even about a deployment model, rather it is really more about a transformation, a paradigm shift and a change in attitude and behavior that is occurring under their very noses."

That's some pretty lofty transformation he's arguing and that shift may well be going on. If it is, I think the transformation is a side effect stemming from how IT delivers services to the organization via cloud regardless of the definition of "cloud" or where it is located. The other side effect is how organizations' expectations of IT are changing.

Still, cloud is most definitely about technology, service models, and deployment models. Maybe my view is narrow because technology, services and deployment are the things I think about. Certainly, the transformation in the data center is shifting from a model of siloed applications and services that are essentially custom-sourced to a generalized model of standardized OSes, networking, storage and configurations defined by runbooks, which are automated in a manner that reduces provisioning time and increases reliability and robustness. This is happening and will continue to happen.

But for it to be transformational, IT needs to buy stuff. When I hear statements like "cloud moves cost from capital to operational," "cloud delivers services without capital cost," "IT won't buy servers any more," "internal cloud computing for enterprises isn't cost effective," and similar statements, I just shake my head in disbelief. In order to deliver these transformational shifts, IT has to spend capital budget.They need to buy servers, software, networking equipment, storage--all of the stuff that makes an agile data center go. IT has to spend operational budget to make the change and it's not necessarily clear that there will be significant operational cost savings in all cases of cloud computing. Regardless, if you stop buying equipment and hardware, then you have to buy services to deliver IT.

I don't believe for one second that there are many organizations that are even contemplating moving all, or even a majority of, their IT applications and services to an external, cloud-based service anytime soon. There may be some small start-ups doing so, but I bet they are the exception and not the rule.  
 
In a way--a very big way---the cloud hype is so far beyond where most organizations are that there is a clear chasm between the two with few clear bridges. IT cares about what cloud is and what it can do for their organization. IT cares about how they will deliver cloud services to the organization, and they care how they will build their own cloud services. IT cares about how they are going to leverage existing and new technologies and products in their organizations. Even something as low level as a networking vendor's preference between Transparent Interconnect of Lots of Links (TRILL) and 802.1aq Shortest Path Bridging will have a profound impact on IT's future product purchases.  

Technology matters. This means products matter, which means capital costs matter. At some point the organization has to select, purchase and implement IT. Some IT services can be outsourced like customer relationship management (CRM) or e-mail, but that doesn't mean all services can or should be outsourced.

The technology and product landscape is changing and it's a really great time to be in IT. But to be able to cut through the hyperbole and see what is really happening, you need to look at the technology that's fueling the transformation under your nose. We've done virtualization. We've done virtualized networking. We've done virtualized storage. We've done runbooks. We've done automation. We've done systems management. These technologies, products, and practices together in a cohesive whole are what we call "cloud," not the other way around.