Ben Franklin said, in describing the challenges of reaching a consensus on the wording of the Declaration of Independence, that it was hard to get “13 clocks to chime at the same time”. That notion fits the challenges of managing mult-vendor projects pretty well too. Most of you know that I started CloudNFV based on the work I did with ExperiaSphere and that I opened my design for use by anyone without restrictions.
I also launched a project with the same name to build an ETSI NFV implementation prototype based on that architecture. 6WIND, Dell, EnterpriseWeb, Overture, and Qosmos all joined in immediately, and Metaswitch, Mellanox, and Schenick all joined later. These companies have all labored to implement the NFV prototype, and all of them deserve both praise for their efforts and a return on their investment. That means CloudNFV has to be productized.
When I stepped down from my role as Chief Architect for CloudNFV in January, I told everyone that this was an essential part of the “Project-to-Product” transition, and it was. I can’t run a commercial activity as an industry analyst. The problem is that while my departure was a necessary condition in the transition, it’s not a sufficient condition. There are many vendors in CloudNFV and each has their own optimum opportunity and preferred path to achieving it. It’s pretty likely that there will be multiple different implementations of all or part of my original architecture that emerge from CloudNFV. It’s also likely that some vendors will elect to do things very differently than the original design mandated, or even assemble partners other than the original ones. All of this could muddy the waters hopelessly, to the detriment of all who participated and all who know about the project and concept of CloudNFV.
To try to “un-muddy” things, what’s now going to happen is that the CloudNFV website will be dedicated to the architecture of CloudNFV, to my original design, and to providing a jumping-off point for those who want to see how the various members of the team have developed the concept commercially. I’ll continue to maintain the website, and CIMI Corporation will continue to hold the Trademark for CloudNFV and the copyright for all the collateral. However, I’ve already delivered a license to the CloudNFV team to allow them to use the term in connection with their past, current, or future activities on the project and also to use the three collateral pieces that were developed for the launch last summer. If the team fields a product or wants to have a project website for CloudNFV, I’ll provide a link on the CloudNFV site so you can find it easily.
My own goal is to take the original design of CloudNFV and expand it to embrace management and orchestration for the cloud, SDN, and NFV and also to embrace legacy network elements seamlessly. I’ve already begun discussions on this with experts in the standards area, and I’m starting operator dialogs at the end of April. This is going to take me some time, simply because I can’t stop doing billable work to volunteer myself as a resource! Hopefully I’ll get some support from outside as I develop the concept enough to prove it’s worth supporting.
I will not be posting material on this new approach to the CloudNFV website. Those companies who committed themselves to the architecture of CloudNFV didn’t commit in any way to my notion of how it should evolve and I have no right to co-opt their work nor to infer their support by posting my new ideas on a site linked to the original concept. I’ll use the ExperiaSphere website for that purpose, and I’ll also provide every one of the CloudNFV project members a link to their own activities that derived from our joint efforts.
Tom Nolle, President CIMI Corporation