SDDC was coined by the former CTO of VMware, Dr. Steve Herrod. The first mention was at Interop 2012. Dr. Herrod was talking about the convergence of networking, storage and server virtualization, and how it would affect engineers and architects and change their vision of the data center.
The idea behind SDDC is fairly sound: Resources become abstracted to run in software to provide consistent, repeatable deployment models. Orchestration and manageability are key. For this idea to provide maximum value, software abstraction must touch everything, not just switches or a SAN or servers. The entire data center construct must be programmable.
While the idea might be sound, many were dismissive of the term because VMware started using SDDC as a marketing tool for its own products, including vCloud Director and its various components, and then Nicira/NSX.
While people may deride SDDC as "marketecture," I see it as something more. It's a push to converge the major components of the data center to reduce complexity and increase the capabilities of the professionals managing those systems.
SDDC doesn't have to be just a marketing term, nor does it only have to refer to VMware. We have numerous options for implementing software abstraction, programmability and orchestration, be it OpenStack with Neutron (nee Quantum) for server/network virtualization or Puppet to automate deployment.
[While technical debates over SDN rage, what will it mean for IT? Check out “Five SDN Benefits Enterprises Should Consider.”]
We also have to consider how SDDC concepts will affect business models. How will your business transform once you can rapidly deploy data center assets with complete visibility into the process? The Phoenix Project showed how devops changed one company's ability to deliver. What is the SDDC version going to look like? How will the organizational structure of a company change when networking, storage and server administration all live under the same roof? And what will it take for organizations to get to a highly automated environment?
Debate over SDDC and its meaning is normal and healthy. The IT industry has done it before with cloud computing and, more recently, SDN. When SDN was first being discussed, many people incorrectly assumed it was only about OpenFlow. Now the definition is evolving to encompass several different technological approaches that have the same general goal: to abstract network resources for faster, more automated deployment.
Skeptics are right to point out the marketing aspects of a term like SDDC, which can become abstract enough to include just about anything. Vendors get accused (often rightly so) of applying flavor-of-the month terms to whatever gear they have laying around.
But over time, industry definitions begin to shake out, and to actually stand for something. I think the same thing will occur with SDDC. I want people focused on solving the problems in the new data center rather than arguing about marketing. When we can have a discussion about the technology powering a software-defined data, then we can really own the term for ourselves.
For those that might be interested in discussing the concept of SDDC a bit further, SDN Central and Gestalt IT are putting on a Software Defined Data Center Symposium in Santa Clara, Calif., on Sept. 10. You can learn more here.
In the meantime, do you think there's value to the SDDC idea, or is it just another acronym that vendors can drape over the same old products? Share your feedback in the comments section below.