Putting the Application in SDN

We would like to highlight a couple of recent articles about SDN that reflect Packet Design’s perspective on the technology. Arthur Cole wrote in Enterprise Networking Planet about SDN in the Enterprise: It’s the Applications, Stupid.” He rightly asserts that the value of SDN isn’t in the architecture itself, but in the applications that the environment supports. It is understandable that during the genesis of a technology, the majority of effort is spent in making it work, but we should not lose site of the fact that optimal application performance is the key to deploying SDN more broadly. And we as an industry are not nearly ready to effectively manage applications across software-defined networks.


IsaacMao via Compfight cc

In fact, Cole cites an article written by our own CTO Cengiz Alaettinoglu in Data Center Knowledge about how traditional, manual management methods are inadequate in a programmable, automated network environment. We need to automate network management best practices and processes to give human operators the visibility and control needed to adequately manage SDN applications in the data center and across the WAN. 

Cole also references a blog post by Mike Fratto (long-time Network Computing editor turned industry analyst for Current Analysis) titled The SDN Application I Want to See. We have not spoken to Mike about the SDN management technology that Packet Design is working on, but his article could be a blueprint for our solution. He discusses why the physical network still matters even with SDN, and that the problem “…is ensuring that SLAs are met in the face of network competition and congestion.” How could the industry do this? Here is his answer: 

“How about this? A network application that monitors network health and takes into account metrics like link utilization, switch queue depths, application flow delay, variation in delay, packet loss and so on, and then maps that data onto a physical topology. Then, when an application is at risk of falling out of its SLA, the network is empowered to move flows around the topology so that the demands of all applications can be satisfied. Applications that are time sensitive like live media go the most consistent and shortest path between peers while bulk traffic is routed out of the way onto longer paths. When that traffic moves, anything that depends on the traffic flows like application performance or security monitoring has to move with it. If network conditions mean that SLAs can’t be met, then alert someone to that fact…” 

This is basically what we are working on with our Network Access Broker (NAB) prototype. Packet Design believes that the SDN architecture should be augmented with a NAB, which would verify if the network can handle the traffic demands of applications without impacting other applications adversely. As Mike says in his post, it would be a complicated system to build. This is because, in a dynamic network, management instrumentation must be collected within the network, not by “peering in” periodically to gather status and utilization metrics. Packet Design is the only company that delivers an always-current model of the Layer 3 network topology – by passively participating in the routed network – which is essential to creating a NAB.

The real-time network provisioning by the SDN controller and the successful monitoring and management of SDN applications requires always-current network models and traffic load profiles. While much of the current industry focus on SDN is in the software-defined data center, where management challenges are less complex, Packet Design is enabling SDN in the routed WAN, which needs best practices and tools for management visibility and policy-based control. Fortunately, current Packet Design technology makes it possible.

Another important element of the NAB, which is key to viable SDN architecture and deployment, is the ability to predict the impact of change to network routing topologies and traffic flows. Unique Route Explorer planning capabilities make it possible to model modifications to the network model and to flow records in real time and analyze their impact.

Cengiz is working on the NAB prototype as we speak, and we plan to have some news about it over the next quarter. Stay tuned for more information!