Infrastructure is often only thought of in terms of the big picture, with little or no attention given to the fact that the whole is only as good as the sum of its parts. That reality is only now beginning to sink in among decision makers as industry efforts to decouple management functionality from the underlying hardware kick into high gear.

Networking in particular has had a tendency of falling through the cracks in the past, but  Mike Harrison, the vice president of Brocade’s IBM unit, sees that changing in the wake of the accelerating shift to software-driven operating paradigms. He dropped by SiliconANGLE’s theCUBE at Big Blue’s recently concluded Edge Summit to share how his company is collaborating with the technology stalwart to fit connectivity into the programmable data center puzzle.

The need to unshackle transport capacity from the physical constraints that have historically complicated management and maintenance is becoming more pressing than ever before as result of the unprecedented disruption happening at the upper layers of enterprise stack, Harrison tells hosts John Furrier and Stu Miniman. He explains that the rapid increase in data from the Internet of Things and the new breed of applications developed in response to that growth are driving demand for both storage and compute capacity.  As the glue that holds it all together, the network is naturally coming under increased pressure as well.

“When you take the next logical step, it’s ‘well, what about provisioning that data from the target device that you just put it on … to get it back to the application demanding it.’ That’s where we come in,” Harrison remarks. “That’s where the infrastructure, from a networking standpoint, becomes critical in terms of what path and selection criteria you use.”

Broadcom is working closely with IBM to keep pace with the fast-changing requirements of customers, he continues, building on a 17-year partnership tracing all the way back to the early days of storage-area networks. The companies are collaborating mainly on integrating their respective solutions so to save customers the trouble of cobbling together the individual components on their own. The goals are the same as always: reduce costs, simplify management and make the network an enabler rather than inhibitor.

“As you develop these faster arrays for storage, as you develop these data-hungry applications and VMs running on denser processors, what you don’t want to have happen is the bottleneck move to the network – certainly not from our perspective,” Harrison notes. Broadcom is trying to address that with new capabilities for managing and monitoring infrastructure, functionality that is becoming increasingly important as users continue to pursue new ways of reducing administrative overhead.

Automation has emerged as a priority in the both the storage world and on the networking side, where it’s taking the form of programmability. The overwhelming majority of organizations don’t have the resources to put together a truly software-defined environment,  Harrison admits, but network functions virtualization (NFV) on the other is becoming more and more practical by the day. The technology is picking up significant steam among cloud and cable providers,  many of which are already harnessing it to streamline service delivery.

Traditional IT outsourcing companies are also adopting NFV software, Harrison says, but for different reasons. Cost is usually the dominant factor. Broadcom is taking every one of its target markets into its account and mapping its roadmap to the specific requirements in each, he concludes, working with partners such as IBM to differentiate from competitors.

photo credit: Daniel*1977 via photopin cc