As NFV spending hits $38 billion, some mobile operators are experiencing teething problems. Vendor lock-in, internal funding issues, management challenges and underwhelming results have dogged many deployments. The following observations are based on 18 months of live NFV trials and deployments as well as insight from ABI Research, Strategy Analytics and JAH Research. They offered their expertise on what NFV strategies sets apart the success stories from the dead ducks.
These are the NFV life lessons taken from trials…and the pitfalls to avoid:
1. It’s complex…embrace it
Whether an operator embraces the appliance approach or prefers to deploy NFV from the ground up, it should expect bumps in the road. NFV takes years or even a decade to implement fully. That means that operators need to be in this for the long game. They’ll need a committed management team and a strategy that’s rock-solid in its vision but nimble enough to adapt to changing circumstances along the way.
AT&T, a genuine global pioneer in NFV, announced that it surpassed its goal to virtualize 30% of its network by the end of 2016. This could be considered a significant success case. AT&T also claims that as this percentage increase, virtualization becomes much more difficult as advanced and highly complex issues need to be resolved. Although AT&T successfully reached its first milestone, it realized that a linear approach to network transformation is not a viable strategy and now relies more on open source and communities to develop open products for its network.
The best thing about AT&T’s approach? Not only is it leveraging others to help achieve its goals, it is also avoiding vendor lock-in and silos.
2. Finance & Operations: on a collision course?
A primary objective for NFV is cost-cutting. However, the financial drivers behind the shift to NFV are, in many instances, hampering its evolution. Operator teams are now reporting that their NFV funding allocation is confusing, drawn out and even incompatible with the goals of virtualization! Categorizing NFV as CAPEX goes against the very nature of a cloud model, which is purely OPEX. Successful NFV deployments involve tough conversations between finance and operations groups. Operators must be prepared to fine-tune their business case processes, CAPEX budget distributions across traditional business units, and OPEX consequences.
3. More haste, less speed
There is much to learn by watching how successful implementations proceed, the steps they take…and the ones they don’t. “Never a new process and a new product at the same time” should be a mantra to live by. Mounting evidence now suggests that CSPs who try simultaneously to deploy brand new NFV functionality and adopt new operations processes for NFV can find themselves stalled by the challenges of simultaneous process and infrastructure change.
Similarly, operators who progress fastest typically move two steps forward, one step sideways. They test the new processes that NFV requires and in parallel introduce new VNF based service functions for one of the few stable services - SD-WAN, vCPE and vEPC - they then bring the new processes and the new software together. This approach leads to greater service agility and faster time to market.
NFV will make significant progress this year. ‘Make haste slowly’ as the ancient Romans used to say!
4. Cooperate to overcome changing landscapes
The transition from fixed appliance infrastructure to virtualized, programmable and distributed architecture presents several technical and business challenges for mobile operators. Decisions taken now on technology can affect the long-term capability to react to a rapidly changing mobile landscape both in demand and mix of protocols. Managing change is a key skill: a case in point is the ongoing, evolving challenge of dealing with the rise and rise of data.
Successful NFV practitioners will evolve and build on the virtualization of the core/CPE with value added revenue services. The more vendors, standards and operators cooperate, the better the chance the industry will realize benefits of elastic and flexible deployment, cost savings (COTS hardware, automation/ zero-touch), dynamic service chaining for personalized services, and time-to-market.
5. A new dawn for subscriber data
Subscriber data is by far a mobile operator’s most valuable asset. Some forward-looking operators are building Subscriber Data Management (SDM) into their NFV infrastructure from the start, putting in a virtual User Data Repository (UDR) as one of the first subscriber-aware applications in the NFV cloud. With this holistic approach, one-to-one provisioning integration is not needed, and subscriber-related information is readily available to applications that want to use it once the vUDR is in place. Operators can quickly benefit from greater agility when introducing subscriber-aware applications.
6. The future of NFV beyond the boxes
Telecoms providers are now moving virtualization beyond the data center to the ‘telco cloud’ and are focused on establishing network-wide NFV management, orchestration and services. CSPs need high-performance service platforms for agile services and best-of-breed VNFs without vendor lock-in.
7. Telecom transformation and NFV
NFV implementation is as much about revamping processes and changing internal mindsets as it is to do with technology. The future of telecom is closely linked to operators’ ability to be operationally agile. Change comes from the top and NFV requires bold top management leadership to move forward.
The lessons summarized here are exclusive extracts from The NFV Playbook 2017, the definitive operator guide to deploying