There was industry chatter that 2015 would be the year of network functions virtualization (NFV). Allow me to disagree: 2015 was not the year of NFV, but rather the beginning of real-world NFV commercial deployments. NFV is not a one-year phenomenon, but rather a journey of network transformation for service providers.
While we are seeing an uptake in deployments this year from major operators, there are many service providers that remain in the education, testing, and trial phases. Service providers are not going to virtualize their networking environments overnight. However, service providers should begin embracing this transformative technology and set forth their initial NFV roadmaps.
Let’s look at what should be considered before migrating to an NFV environment.
Service providers must first develop clear strategies and goals for their NFV deployments. Most network operators should think about NFV in the context of a specific use case and business driver that will provide the highest return. For instance, a common business goal is achieving operational efficiency by deploying business VPN services using vCPE or virtual managed services. Other providers are simply looking at automating their network device provisioning services.
Depending on which NFV deployment a service provider is planning, before rolling out a specific virtual function or automating a new or existing service it’s important to evaluate current processes, skills, and operational models that may need to be revamped. In most cases a hybrid approach of physical and virtual functions can achieve the overall business outcome.
Misconceptions About NFV
Since the inception of NFV, the industry has been discussing its purported benefits – increased agility and speed to market, reduced equipment costs, reduced power consumption, and so forth. Service providers deploying NFV today are reaping these benefits and entering new markets they haven’t addressed before.
But keep in mind: While NFV is often beneficial, service providers should not virtualize a given function simply because they can. A provider should consider what specific business outcome it seeks to achieve – scale, agility, performance, economics? Or a combination of these benefits?
Service providers should also not continue to follow the same operational approaches they are used to, overlooking the functional benefits NFV brings to the game. In today’s environment, most services offered are tied to underlying physical devices, and the associated OSS is built around those silos and based on an inventory of those physical devices. The NFV environment is virtual, and by the nature of being virtual, there is no concept of inventory as such. So an operator will need to re-think how to operationalize its services from the creation, provisioning, and assurance perspectives – and may need to rethink its operational systems for NFV.
In closing, by working closely with their NFV technology suppliers, service providers can begin to simplify their service creation and deployment processes with orchestration and automation capabilities that accelerate time to revenue. We are clearly still in the early stages of a complete transition to an NFV and SDN environment. While there are still a few challenges to address, it is great to see the progress the industry has made toward a virtualized environment. I look forward to seeing more commercial deployments in the coming year and sharing the lessons learned.