The Open Networking Summit currently taking place in Santa Clara, Calif. is like attending a giant science fair for the networking industry. There are arcane demonstrations detailing how software-defined networks and the OpenFlow protocol will change the way networks are built, managed and operated. There are speakers from Google, Verizon and Yahoo detailing their projects and successes with OpenFlow as well as investors and bankers swarming the whole event. But there isn’t an app store.
In several sessions this afternoon a focus on programmability and creating networks that are developer-friendly showed how far networking has to go and how revolutionary a change the industry is undergoing. Networking, be it in a data center or inside your home, is facing a glut of traffic that is moving in ways the original network architects could have never predicted. Thus, it is becoming more complicated, slower and more costly.
The creation of the OpenFlow protocol, which separates the act of directing how packets move across a network from the physical act of moving those packets, has helped create excitement around networking, and is precipitating change. The change is actually the creation of software-defined networks that are programmable (for the record, a software defined network doesn’t need OpenFlow). There’s also a third change that’s been going on regarding the commoditization of networking hardware and the rise of merchant silicon, but that’s a side issue.
But since the promise of SDNs is really the promise that engineers can program new services quickly and easily on the network without having to worry about the underlying infrastructure, the question is how easy writing, finding and running such programs will be. And will the enabling software layer be open or closed? In short, will OpenFlow be the Android (s goog) of networking; in that it brings an open system to hardware vendors that still adds value to those manufacturers and creates a new layer of value for app developers?
Shehzad Merchant of Extreme Networks (s extr) asked this question in his talk when he laid out some of the things the industry still needs going forward. Among the list was better hardware support at the silicon level, advancement of the OpenFlow protocol, and for the folks at the Open Networking Foundation, who are promoting OpenFlow to create an app store. Greg Papadopoulos at NEA also emphasized that most of the value created by this change in networking will accrue to the providers who create the most attractive platform for developers.
Just because a company buys software (or gear) to install a programmable network doesn’t mean that it wants to hire the engineers and talent who are able to program it. A far better solution is to buy applications that will run on the controller or maybe at a layer above the controller. But right now it’s not clear if a universal and open controller will dominate the industry. There’s Floodlight, an open source controller that BigSwitch is pushing, but there are also other efforts, such as the controller software from Nicira, which is using OpenFlow but is essentially closed.
If everyone goes the Nicria route, then network engineers will have to write their apps to a specific controller as opposed to writing apps that they could sell across the industry. It’s the difference between writing an app for a few handsets running Apple’s iOS (albeit wildly popular ones) or writing for a slew of them by developing on Android. There’s no guarantee that this is the right way to view the industry this early on in the game, but it’s a lens worth trying as the ecosystem is shaping up.