Nearly nine out of 10 North American enterprises will have software-defined networking (SDN) live in their infrastructures by 2016, according to a recent study by Infonetics Research. With that level of adoption forecasted, no matter the technologies or vendors in your SDN stack, or whether the underlying resources are on premise or in public clouds, now is the time to start thinking about how the shift to SDN can affect day-to-day performance monitoring of your infrastructure.
It's clear that performance monitoring technology must evolve to keep-up with SDN's dynamic nature. Real-time service provisioning and topology changes need to be reflected within the system immediately, or IT pros will suffer significant gaps in network performance visibility. In fact, I'm seeing three major impact areas where visibility gaps could result if companies don't effectively monitor the IT infrastructures created using SDN technology.
Impact of dynamic, real-time change
The promise of SDN changes the infrastructure paradigm: Gone are the days of static configurations of network devices. With SDN, new network devices and features can immediately become available for use. If your performance monitoring solution is running only daily checks on what's new in your environment, this dynamic, real-time change of configurations can create significant performance visibility gaps. What's needed is a performance monitoring solution designed with APIs integrated within the SDN systems, listening to the network for the addition of new devices (physical and logical), segment capacity increases, modified traffic patterns, and more.
Impact of rapid, on-demand growth
As creating, modifying, and tearing down networks via SDN systems becomes a trusted IT process, network capacity, much like today's automated compute capacity, will evolve to flex up and down based on demand. As network capacity is added, performance monitoring products will need to understand and immediately adapt to SDN's changing physical and virtual connections.
However, meeting this demand isn't as simple as adding the new network devices and segments into a performance monitoring solution in the traditional way, as we described above. Similar to how SDN systems add new network capacity, performance management systems must be able to add monitoring capacity to meet the rapid growth needs of the network, and continue analyzing business service performance.
Impact of customer service
When an SDN system and performance monitoring solution are integrated within a customer service context, an entirely new set of use cases can be supported. For example, instead of performance monitoring software listening only for new devices to be added into inventory, the monitoring solution will recognize the role each new device plays in supporting a particular customer or tenant of the network.
In this scenario, users of the performance monitoring software not only know about the health and performance of individual devices or links on the network, but they can also answer questions like, "How is the performance of the HD videoconferencing system we sold to a Boston customer affected by network changes, and what action must we take to maintain SLAs?" This type of shared context extends monitoring to the service topology, meaning both SDN and performance monitoring share the knowledge of physical and logical connectivity of the devices (both physical and virtual) that make up a service, both in real-time and historically.
Let's take this a step further. Imagine an SDN system proactively querying performance monitoring software about the health of a particular customer's service and how that service might be affected by network changes, even before the changes are made. Based on the performance data, the SDN system could take preemptive action to ensure network resources continue to meet customer SLAs. This type of continual optimization leverages shared customer service context to create a feedback loop between SDN and performance monitoring.
The good news? It's still early in the evolution of defining and deploying SDN. However, as SDN makes its way from lab trials to production, we need to keep in mind how dynamic real-time change, rapid on-demand growth and service context will play a key role in enabling a successful SDN deployment and avoiding performance visibility gaps in the IT infrastructure.