Where does xTCA fit in SDN?
It’s widely known throughout the carrier world that there is an alarming trend in the ARPU, as well as cost per bit. This has driven a significant interest in the carrier world in Software-Defined Networking (SDN), which separates the control and data planes in network switches. At the first day of AdvancedTCA Summit 2012, many of the presentations – and much of the hall talk – was centered around OpenFlow, the preeminent SDN specification, and how it can/may be used in future carrier network architectures.
On Day One of the Summit, Dr. Stuart Elby, Vice President, Network Architecture and Technology, Verizon Digital Media Services, confronted this topic in his keynote “How OpenFlow Can Revolutionize the Carrier Business” by explaining how dynamic, service-aware networks based on OpenFlow allow for quick introduction of new services and reduce cost. He also, however, pointed out several of the factors that are limiting the progress and adoption of OpenFlow at present, which included:
• The split between networks and the data center
• OpenFlow’s focus on only the lower-layers of the network
• OpenFlow’s lack of transport technology beyond Layer 2 and Layer 3 (for example, optical transport)
• The lack of a system ensuring stable, robust operation in a production environment
• The lack of a migration roadmap due to the limited availability of hybrid OpenFlow switches and routers
After reviewing these shortcomings, Elby posed a few questions to the AdvancedTCA crowd regarding what the xTCA ecosystem had to offer in terms of helping carriers leverage OpenFlow (initially in niche applications, and eventually on larger scales) and optimizing the dynamic management of the virtualized data center and network resources. More specifically:
- Can AdvancedTCA solve the OpenFlow hybrid switch problem?
- Can AdvancedTCA help with the integration of services within OpenFlow network elements?
- Can AdvancedTCA help with the multi-layer integration of OpenFlow
Because SDN is enabled by advances in compute technology, in closing Dr. Elby asserted that “there is an opportunity for ATCA-based systems to close gaps and accelerate OpenFlow deployments.” I look forward to your thoughts and comments on these questions, as well as further discussion of the possibilities of xTCA in OpenFlow. We know Dr. Elby does.