Real-time Metrics Are Critical to NFV Deployments

CSPs will deploy new services faster than ever before, efficient use of network resources will become the expected norm versus the hoped-for goal, and softwarebased network functions will be more widespread. Yet, as we contemplate this exciting chapter in the telecoms industry, we must realise that although NFV will usher in many efficiencies, none of this new virtual reality will operate correctly without the right metrics. Successful NFV performance needs real-time network and customer intelligence with the granularity to make it actionable, profitable and capable of ensuring an outstanding customer experience.

Capturing and calculating real-time metrics requires service assurance tools that possess the agility and scalability to meet the demands of dynamic NFV environments. The best NFV service assurance solution performs the processing of IP packet data locally, on software-based probes placed near the point of traffic capture, to ensure the integrity and timeliness of the data. There are four important aspects of NFV that depend heavily upon the availability of real-time metrics: proactive problem detection and resolution, just in-time capacity management, closed-loop orchestration and delivery of new and differentiated services.

Automated attention to detail

Analysis of real-time data generated by NFV-based networks has the power to change customer perceptions and experiences for the better. Up-to-theminute usage and performance data allows CSPs to proactively monitor subscriber network and service experience, while NFV-enabled orchestration reallocates resources based upon traffic demand. Monitoring and real-time analysis of performance metrics help to identify troublesome network elements and ensure that policy settings generate corrective actions – before these problems affect customers. This automated attention to detail creates networks that consistently provide positive experiences, which makes for happy, loyal customers and reduced churn.

Today’s networks rely upon tightly integrated services software and purpose-built hardware. In addition, network resource planning is plagued by long scheduling and procurement cycles in which equipment and software are ordered and paid for months in advance, depleting capex and monopolising time. Capacity forecasting is not an exact science and often involves ordering new equipment before it is actually needed. Sometimes network element capacity limits are set as low as 40-50%, just to allow for the long procurement and delivery lead times. A distributed, software-based architecture running on commercial off-the-shelf (COTS) hardware and powered by realtime metrics, eliminates forecasting errors and decreases unnecessary capex spending. CSPs can set higher threshold triggers – around 70-80% –before just-in-time capacity is automatically deployed and a new server is provisioned, exactly where and when it is required.

NFV gives CSPs the ability to place virtual network functions (VNFs) hosted on COTS hardware, anywhere in the infrastructure and to make adjustments to these VNFs on the fly. Software-based virtual networks are intended to be self-managing, so real-time network data must be fed into the service orchestrator, which enables the network to become more self-regulating and service-aware. This dynamic orchestration removes the need for manual adjustment, which equates to reduced opex and decreased potential for human error. Automatic bandwidth adjustments are made in response to shifts in traffic and network congestion, ensuring that capacity is neither over-utilised nor under-utilised and that network and service resources are available as needed.

Evolving preferences have caused customers to expect new services that are delivered on-demand. As a result, CSPs must be much faster to market with competitive, new services. NFV-enabled networks provide the agility required for CSPs to quickly and efficiently deploy new services, with automated bandwidth adjustments made based upon service traffic needs. This powerful service agility depends upon real-time views of IP packets which allow CSPs to actively monitor service uptake and performance. Like pearls in an oyster shell, IP packets contain priceless bits of intelligence that supply critical insights into service and usage experience. Without real-time visibility to IP packet performance metrics, how can CSPs know whether new services are working properly, what kind of experience users are having or if network resources are consistently utilised?

New revenues

The evolution to a service-agile infrastructure opens up brand-new revenue opportunities, especially with small and medium-sized businesses (SMBs), a growing segment previously too challenging for most CSPs to serve effectively. Savvy CSPs will position themselves to SMBs as leaders with the technological acumen to provide innovative services, educate them on new technologies and engage them in partnerships that create new business initiatives. Real-time data from virtual customer premise equipment (vCPE) allows CSPs to quickly respond to the changing requirements of SMBs, deliver customised resources and to insert services such as security and configuration settings into the network, relieving SMBs of this technical responsibility.

Preserving the availability, reliability and timeliness of critical network and service data that drives NFV efforts, is not a solo endeavour. CSPs must choose a service assurance partner offering tools capable of not only providing real-time data capture and correlation, but a partner experienced in testing, monitoring and leveraging NFV-enabled networks. Collaboration with a qualified, knowledgeable service assurance partner is essential to guaranteeing the just-in-time capacity management, efficient service orchestration, cost savings and revenue potential associated with NFV.

To contact an NFV Service Assurance expert or to get more information about NETSCOUT’s NFV Service Assurance solution, please visit http://www.tekcomms.com/virtualization.