A data center might look like a room full of black boxes, but it’s obviously a busy place. Every second, millions of packets come in from around the world, get moved through switches, and then sent back out at just about the speed of light. So it’s no wonder software-defined networking — an emerging networking architecture that speeds up switches, makes management easier, and gives IT managers more control over the network — is gaining in popularity.
So what exactly is software-defined networking?
Think of the data center as Grand Central Station in New York. If you were to arrive from Chicago, and you wanted to go to Boston, you’d need to switch trains. But there are dozens of ways through Grand Central that can get you there — what’s the fastest one? There might be a big human traffic jam around the subway entrance, or construction could be blocking access through the food court.
Without knowing about these conditions, you could get delayed very easily. But if you had an all-powerful, all-seeing pedestrian traffic cop who knew everyone’s destination and could tell them the best route to get there, everyone would breeze right through.
Software-defined networking is essentially a software-based “traffic cop” for all the packets that come through your customers’ switches. But what separates it from traditional, hardware-based “thinking” switches is that it uses software — which is much faster than physical switches — to do the thinking. At the click of a button, IT managers can define protocols on the basis of availability, efficiency, or capacity.
Here’s a concrete example of how it works: Let’s say an online retailer gets a huge traffic spike during the holidays. Gene, the retailer’s IT manager, could use software such as OpenFlow to route packets to a cloud provider once his own data center reaches a certain capacity. After the holidays, Gene can then define a different protocol, quickly and easily, without having to go in and manually configure the switch.
Sounds great, but how do I build a bigger offer around it?
With the rise of Big Data, virtualization, and BYOD (bring your own device), more and more data centers will require these faster switches to improve their customers’ experience, optimize performance, and ensure network availability. But what your customers may not realize is that software-defined networking places a huge responsibility on your customers’ data centers:
How do your customers optimize performance and protect uptime when complexity meets criticality?
There are two clear solutions:
- Modular, high-performance power and cooling systems: Traditional power and cooling systems aren’t agile enough to support such elastic virtual switches. Your customers will need equipment that can respond to the dynamics of a software-defined environment. And don’t forget: the more switching, the more business processes they have online, the more critical their data centers become.
- Data center infrastructure management (DCIM): Your customers will miss out on the primary benefits of implementing SDN if they don’t combine it with DCIM software. Why? Because in a software-defined environment, the only way to ensure packets are sent to an optimal rack is to have visibility into the physical infrastructure. DCIM provides one-pane-of-glass visibility across your customers’ entire data center, and is therefore able to inform the SDN software where to send packets for optimal efficiency and availability. Without DCIM, SDN creates a more complex and critical environment without a way to manage it properly, putting your customers’ data centers at risk for downtime and stranded capacity.
SDN is trickling down from the enterprise to the smaller data centers, and it’s only a matter of time before it’s as common as server virtualization. What you can start doing now is prepare for its emergence, and these two strategies help you do just that.