YOU ARE AT:Big Data AnalyticsWhere Hadoop fits into software-defined networks and virtualization

Where Hadoop fits into software-defined networks and virtualization

Hadoop and other big data analytics platforms have drastically changed the manner in which companies view their information repositories. Now, organizational leaders view every piece of data – no matter how seemingly insignificant it may be – as a piece of a larger puzzle. When large volumes of information are processed through advanced analytics systems, businesses can retrieve new insights that help them enhance their operations in numerous ways. For telecoms, Hadoop presents an opportunity to better engage clients, more effectively market their services and identify opportunities for business growth.

One of the concerns with big data tools, however, is the technology's total cost of ownership. The amount of processing power required to properly run a Hadoop-based analytics solution can be substantial. Not all telecom networks are currently configured to handle such a stressful workload. By employing software-defined networking tools, carriers can better utilize their network resources to power Hadoop operations.

SDN supports Hadoop applications
Telecom veteran and Network Computing contributor Richard Kagan highlighted an experiment conducted by Infoblox, which found that Hadoop could more efficiently process datastreams when operating across an SDN-enabled network. The application was able to utilize the dynamic flexibility of SDN to prioritize certain traffic ahead of others depending on the HTTP ports they originated from. This capability allowed researchers to process important data first above other traffic types, resulting in a quicker and more efficient use of the analytics tool.

"Applications like Hadoop rely on distributed computing in order to achieve acceptable performance," Kagan wrote. "Distributed computing requires a network. If SDNs can enable applications to make better use of distributed processing resources, then big data applications can deliver acceptable performance using much less computing hardware, making them affordable for more organizations."

Researchers from Clemson University conducted similar experiments to test the effectiveness of Hadoop when run across an SDN-enabled network. There findings concluded that SDN technology allowed users to view data traffic in real time and make changes as necessary. In addition, the research team discovered that they were able to effectively monitor the flow of information, giving them greater oversight across the massive number of nodes needed to store big data volumes.

As telecoms increasingly look to data analytics as a means to enhance their operations, they may need to build out their network capabilities to handle the rigors big data frameworks such as Hadoop. SDN not only presents an opportunity to enhance and accelerate customer-facing services, but to improve carrier analytics projects as well.

ABOUT AUTHOR