Proprietary networking has always been a double-edged sword. It delivered the stability businesses depend on to keep things running, but it often lacked the innovation that help businesses get ahead. More to the point, most new technology introductions were taken with skepticism and caution. Each new generation of products only seemed to lock businesses in even further with the traditional vendors like Cisco Systems or Juniper Networks, delivering new capability but with a hook—a very expensive hook.
Open source solutions, on the other hand, appear to be the antithesis of proprietary networking: plenty of innovation, an ability to move fast and dramatically lower cost. What’s not to love? Well, for starters, choice. Nobody likes a monopoly, but on the other end of the spectrum, having too many choices makes the evaluation and selection process nearly impossible. The business world needs a better alternative, and the open source community is listening, beginning to deliver this in a more cohesive manner.
The historical problem with open source was that, while better alternatives existed, the burden of integration landed on the customer or a specialized integrator (who simply created a different kind of lock-in). Most businesses tended to stick with a proprietary solution because it already worked with what they had and there was lower risk if things went wrong. IT was hesitant to open that can of worms, and business leaders weren’t interested in rocking the boat. This was until the proprietary solutions limited a business’ ability to change or innovate, which puts new pressure on IT to come up with a better solution.
To be fair, all enterprise solutions will, by default, create a lock-in as once they are deployed they are very difficult to replace. The benefit of open source is not that it removes lock-in altogether; instead of being wed to single vendor and a single technology, one is connected to a platform / API. While the open source solution may be driven by a single organization, there are a wider range of technology choices underneath.
Two large open source organizations, the Open Networking Foundation and ON.LAB are merging, signaling that they understand the business need for a more fully baked open source solution instead of just projects. The idea that “If we build it, they will come” is stepping aside, being replaced by the realization that “We can build it, but if we don’t make it easy and interoperable, they won’t come.”
With the merger of these two organizations, the companies and individuals involved are now contributing to projects that are more cohesive and more “commercially adaptable”, enabling greater reach than before.
The combined organizations are working to create an “innovation pipeline” to help overcome the biggest challenge with open source acceptance: the need to string all of these pieces together yourself. By shifting the focus from projects to a pipeline, they can establish a stack that enables multiple projects or platforms to be easily chained together, working to deliver the services that businesses need. This chain is comprised of both the solution blocks (projects and platforms) as well as the “glue” (the standards and internetworking APIs that link them all together). If one envisions this like Lego blocks, the blocks themselves are the projects and the small, standardized round pegs on the top are the APIs, allowing pieces to be combined into a solution. By standardizing the connections and interfaces between these blocks, more interesting and customized solutions can be built because less time needs to be spent on the integration.
An example of how this pipeline works is CORD, a project to re-architect a telecom central office as a datacenter, enabling a more flexible and distributed solution. While this is a defined platform, variants for mobile (M-CORD), residential (R-CORD) and enterprise (E-CORD) can easily be designed by “snapping” in different pieces, like building different models with Legos, using many of the same base blocks but some specialized blocks to differentiate between a car and a house.
The beauty of this approach is that it enables other open source alternatives to be integrated into the design, for instance adding alternative frameworks like OPNFV or alterative controllers like OpenDaylight, without requiring someone to re-architect everything from top to bottom.
While much of the early work will be focused with telecom operators and service providers, that makes the most sense because they are the early adopters these days. The old stodgy phone company has become the innovator as carriers reinvent themselves, shedding years of proprietary equipment in favor of more flexible, open solutions. We’ll see this trend play out with enterprises as well. They are behind carriers on the open platform revolution but can be the fast followers as they see success in business transformation taking hold with carriers.
While this innovation pipeline will not replace proprietary networking, it does put more pressure on those traditional networking vendors, which is a good thing. That is how markets are supposed to work. Competition for ideas drives innovation, and in the end, business are the ultimate beneficiaries.