There is no doubt that the current race among CSPs to be first to launch the most innovative network as a service (NaaS) offerings is disrupting our industry, allowing us to participate in the most exciting technology evolution in a decade or more. The transition from traditional network architectures to those built on virtualized network functions and network controllers will be a long journey, but it is one that promises a windfall of benefits for CSPs, MNOs and their customers.
Network functions virtualization (NFV) and software-defined networking (SDN) will reshape the networking landscape by making networks more dynamic, flexible, resilient and scalable. These networks will be, in the vision of the Metro Ethernet Forum (MEF), agile, assured and orchestrated. In these environments, CSPs will launch innovative services and support on-demand requests from customers.
Despite the appeal of NFV and SDN, their widespread and immediate adoption still face an uphill battle, because network technologies will become increasingly complex and NaaS offers will need a new generation of network architectures and OSS platforms. At the end of the day, even if CSP customers find benefits in new technology, they primarily want reliable services, at the right price, supported by strong SLAs. New technology cannot be adopted at the expense of the quality of service (QoS) delivered by their business partners, and if customers do notice a disruption in service, whether it’s problems with speed, reliability or lack of agility, they could seek to renegotiate their contracts as soon as possible, or possibly turn to a competitor.
To avoid these circumstances, CSPs need to begin preparing now, by taking certain measures and trialing the technology needed to transition to this new virtualized environment.
A New Service Environment
The first step for innovative CSPs transitioning to an NFV– or SDN-enabled environment is to put in place modern service level management (SLM) solutions to assure customers’ service levels. Once these solutions have been adopted and CSPs have adopted a new, future-proofed, unified end-to-end approach to service assurance and network performance, the next era of increasing network agility and configuration automation can begin.
With comprehensive performance visibility, CSPs’ operations teams will be able to monitor, measure, discover, correlate and visualize tomorrow’s networks in parallel with existing WAN infrastructure. This allows them to deliver high QoS and fulfill customers’ SLAs, while migrating them to advanced networking technologies.
And, despite the difficulty of monitoring dynamic networks that comprise both physical and virtual elements, CSPs have been meeting this service assurance challenge. As the level of network automation continues to increase, CSPs can improve their management processes, smoothly transition to virtualized environments and evolve their network infrastructure by following these five pragmatic steps:
1. Managing virtualized network functions (VNFs)
Since virtualization obstructs CSPs’ network visibility, the only way to see how these functions are actually performing is through performance assurance statistics, made available in VNFs by accessing devices directly or through vendor element management systems (EMS). As VNFs are rolled out, CSPs must add real-time, on-demand performance-monitoring capabilities to their OSS and make sure their operations team’s network management toolset is agnostic from the virtualization layers. At the end of the day, SOC and NOC engineers should not be concerned about a network function being virtual or physical. Instead, they should be provided with solutions that allow them to understand the performance of each network function supporting a given customer’s or subscriber’s service.
2. Troubleshooting network function virtualization infrastructure (NFVI)
As CSPs migrate physical network functions to a multitude of distributed and specialized NFVI, engineering teams will need to troubleshoot the underlying telco cloud infrastructure introduced in their networks. And, as with every distributed and multi-vendor environment, this will require a higher level of troubleshooting capabilities. In addition, CSPs should ensure that orchestrators are publishing inventory information about the network functions they are hosting, as well as statistics about the physical architecture that supports those functions.
3. Scaling VNFs
As the number of VNFs grow and network agility increases, capacity reporting and NFV framework capacity management capabilities will become critical. Virtualization places immense stress on the back end of a network infrastructure because, as the network evolves, new and old systems must integrate together seamlessly. The pieces won’t always magically fit together – they need to be properly monitored and managed as VNFs scale.
4. Enabling “Networking as a Service”
Many network controllers are focusing on automating device configuration, but reliable NaaS will only be made possible when controllers implement a feedback loop and leverage end-to-end OAM metrics as part of their decision-making process. After gaining access to these performance metrics, CSPs can make more informed decisions to ensure that end-to-end service quality is not impacted by configuration changes dictated by siloed and automated optimizations process.
5. Assuring SDN and WAN infrastructure
Today, most management networks are not fully assured – that needs to change in order to ensure that the entire telecom infrastructure operates properly. As an example, a network controller requires a control plane to send commands to equipment. By assuring availability of all NaaS technology introduced in their networks, CSPs can ensure that new NaaS architecture and all its components are performing and able to support future service growth.
The Path to Success
To maintain a consistently high QoS for customers in the face of growing virtualization, unified service performance assurance is critical. While the migration to NFV- and SDN-based services seems daunting, CSPs that take the time to upgrade their service assurance solutions will allow their operation and engineer teams to successfully manage these functions and make the most of the current transition. From the rollout and enablement stages of virtualization through to the management of an entire SDN architecture, service assurance functions are evolving to be part of the strategic arsenal of CSPs.