An OpenFlow-like approach to networking was not really possible until recently, not just because the idea is novel, but because network hardware was not capable enough to handle traffic in this way. Now that we're seeing switching platforms such as Intel's Seacliff Trail in production, we have the means to make OpenFlow functional, fast, and scalable.
There are switches and other network devices that support OpenFlow now, but they're generally built as traditional devices and run their own operating systems. They support OpenFlow, but are not strictly OpenFlow devices, and thus have little to no cost savings associated. When the OS is removed and the switch becomes a white box, we're suddenly dealing with hardware, and the costs should drop dramatically, along with the associated bits like optics.
OpenFlow has a long road ahead, however -- a road littered with potholes such as massive established networking vendors who aren't keen on losing their high margins and hefty market share. There's also a natural evolution of any disruptive technology, and OpenFlow is still in its infancy. There's a lot of work to be done developing tools and firming up standards before the worm turns on the traditional network, but smaller companies and hardware manufacturers see a market opening up, and they are ready to take advantage of it. To paraphrase Stu Bailey, anyone playing in the networking space over the next few years will have to make a decision whether to sell software to run on hardware from any manufacturer, or sell hardware to run software from any developer.
What's clear is that with OpenFlow, we are ultimately going to see the cost of networking drop while capabilities and features increase. That's good news ... for most of us.
This story, "Why OpenFlow is the next big thing," was originally published at InfoWorld.com. Read more of Paul Venezia's The Deep End blog at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.