Reality or illusion? Challenges of NFV adoption in the enterprise

To further strengthen the use case for enterprise NFV, a recent market research report from ACG Research shows that NFV adoption will help keep the total cost of ownership (TCO) of physical equipment below 62%.

However, industry surveys and analysis reports show that less than 15% of enterprises have deployed some NFV in their production environments. On the other hand, more than 50% are exploring their choices in NFV strategy and vendors. The key question is: Since NFV provides a lot of potential upside for IT business, why don't enterprises actively adopt NFV? The following will analyze three major reasons:

• Organizational disruption, skills shortages, and associated risks
• Performance (SLA) and security (compliance) challenges
• Uncertainty (market maturity) and unpredictable ROI

In fact, while the ultimate vision of NFV and software-defined networking (SDN) is very attractive and desirable, the potential pitfalls and associated consequences make enterprises cautious about the adoption of the technology. So we will analyze these unfavorable factors to better understand these reasons and discuss how to help enterprises get rid of the barriers of technology application and move towards NFV.

[[207464]]

Organizational disruption and skills loss

The network team is the one that knows the network best. They are responsible for the operation of the entire network. The same is true for application, security, server, and virtualization teams. NFV spans all of these areas, so there is a risk that the organization will operate poorly.

NFV also places a skill demand on existing teams, including selecting the right servers and server configurations, hypervisors, resource allocation, and service linking. Without the necessary knowledge and skills, NFV initiatives will fail.

The key to avoiding these risks is to adopt an NFV platform that IT teams can purchase and deploy without the need for advanced virtualization expertise or help from a virtualization team. Emerging NFV platforms provide hypervisors, CPU pinning, non-uniform memory access (NUMA) boundary setting, single root input/output virtualization (SR-IOV), drivers, physical and virtual port mapping, and many other services.

These platforms enable streamlined creation of services between virtual functions, again eliminating the need for specialized skills, allowing businesses to become software-centric in the short term with minimal operational or organizational disruption.

Performance and security challenges

Many enterprise applications are business-critical applications with high traffic, complex configurations, and strict requirements for compliance and end-user experience. The expected benefits of deploying NFV, such as reducing the time required to deploy services or becoming more flexible and efficient in using IT infrastructure, should not outweigh the business costs of downloading, executing, or being affected by the application.

Because servers were originally designed for application workloads, rather than network and security workloads, general-purpose hardware, hypervisor overhead, VM contention, and virtual switches can all interfere with the performance level agreements (SLAs) required to connect enterprise applications to meet and maintain the necessary service levels. Additionally, shared virtual environments inherently conflict with the isolation required to meet compliance requirements.

To minimize the impact of these challenges, new NFV platforms reserve and allocate hardware resources for each virtual function, such as CPU cores, hardware-accelerated SSL, memory, physical and virtual interfaces, etc. The result is a combination of the flexibility of cloud and virtualization with the performance of dedicated hardware devices, and this approach also provides visual isolation of independent operating systems with partition management to fully address the disadvantages related to performance and security.

Uncertainty and unquantifiable ROI

One of the biggest issues regarding NFV is uncertainty and the need to establish a reliable return on investment (ROI). This challenge arises because the goals of increasing agility and reducing capital expenditures and operating costs are vague when not tied to a specific implementation or use case.

One of the best ways to mitigate this risk is to look for NFV platforms and approaches that deliver proven and needed capabilities while laying the foundation for broader NFV adoption in the future.

If the enterprise decides not to pursue a broader NFV strategy, or takes a different approach to implementing NFV, the ROI on the initial NFV platform purchase has already been realized by supporting needed and mature network and security functions. If the enterprise adopts a strategic NFV focus and finds that they are already on solid foundation, more ROI will be realized.

in conclusion

The key to driving enterprise adoption of NFV is to uncover unknowns, address disadvantages, select use cases, and find ways to achieve NFV goals while avoiding risks and potential pitfalls. By looking for emerging platforms that address organizational disruption, skill gaps, performance and security challenges, and the need for a proven ROI, at least 50% of enterprises currently exploring their NFV options will be able to stand out and transform to SDN and NFV.