Network operators envision a future where new services are conceived, designed, and deployed in days or weeks – as opposed to months or years, as is typical today. To realize this vision, operators are transforming their legacy networks with network functions virtualization (NFV). With its promise of faster innovation, improved agility and cost savings, NFV simplifies configuration and control of networks and reduces the time it takes to rollout and assure new services. NFV is viewed as the critical key to enable network operators to compete with cloud-based over-the-top (OTT) providers in coming years.
But achieving the benefits of NFV goes beyond a simple network transformation – it also requires a business transformation. Today, most network operators have a highly-fragmented approach to designing and deploying new services, requiring many different systems and methodologies to test and assure service quality and network performance at each stage in the service lifecycle. This siloed approach is one of the biggest obstacles to realizing NFV’s cost and performance benefits.
Legacy Silos Causes Lengthy Delays
Traditionally, systems for testing and assuring new services and underlying network functions are divided by organizational group, network domain, technology, and data source. As a result, new services and network functions are verified in isolation, and then manually handed off at different stages of readiness. For example, extensive pre-launch testing is typically performed in labs by multiple development teams, and then, when the service is ready to go live, another set of testing and assurance systems are created by operations teams to turn-up, monitor, and troubleshoot services.
Operational assurance systems are often deployed months after the initial service launch and have no linkage to the testing tools or methodologies developed in the lab. Likewise, as operational issues are identified, and assurance systems are updated, feedback to lab-based testing procedures is very limited due to a lack of consistency in test methodologies. This manual and tedious process can lengthen the launch cycle of new services to months or even years.
The Need for Lifecycle Service Assurance
By incorporating a Lifecycle Service Assurance (LSA) approach, these operational challenges can be transformed into opportunities. LSA is based on DevOps principles widely adopted by software enterprises to streamline new product releases by breaking down barriers between development and operations teams. LSA extends these principles to network operators with a focus on the unique challenges of automating testing and assurance in communications networks.
One of the unique challenges network operators face stems from the expectation that NFV will be adopted gradually over the next decade, as operators have invested billions into their existing legacy infrastructure. LSA specifically addresses this challenge by streamlining testing and assurance processes across physical and virtual (NFV) environments, as both will likely coexist for many years.
The Importance of Active Testing
While NFV brings challenges, it also creates exciting opportunities. One example is active testing of services, which delivers compelling benefits for certain use cases such as:
• Verifying a new network element or service immediately after turn-up
• Monitoring SLAs even when a service or network function isn’t being used (e.g. for an alarm service)
• Segmenting the specific location, function or protocol layer responsible for an SLA violation
Unfortunately, active testing has historically required deployment of complex physical probes into the network. Due to the expense of the physical equipment, these probes were typically deployed in a limited set of network locations for applications such as Ethernet turn-up verification and SLA monitoring. With the shift to NFV, all network functions run on common compute, storage, and networking resources. This enables software-based virtual test agents (VTAs) to be deployed on-demand, anywhere in the network.
A key element of LSA is the use of virtual test agents to align methodologies and reuse code across lifecycle testing needs – thereby transcending the barriers between development and operations teams. LSA also embraces the ETSI NFV active monitoring framework, which enables automation of testing and assurance processes via integration with NFV management and network orchestration (MANO) functions.
Getting to zero-touch automation
DevOps principles include the concept of automating testing across the lifecycle of a software release. In complex, dynamic NFV networks, simple, script-based automation can’t fully automate problem detection and resolution processes. To achieve true, zero-touch automation, network operators need to move beyond simple automation to intelligent automation.
One of the most valuable components of LSA is its ability to automate operational assurance by leveraging machine learning and artificial intelligence based on predictive and prescriptive analytics.
With an LSA approach, predictive and prescriptive analytics trigger automatic decision-making to anticipate issues before they appear and apply remediation to resolve the problem.
Examples of automatic, zero-touch automation responses include:
• load sharing to remedy an overloaded network function
• re-routing of services to overcome a network failure
• restoration of virtual network functions to improve poor service quality
The result is an agile network that becomes more efficient over time, as it continuously improves and self-heals.
The bottom line: NFV requires a cultural shift
NFV technology makes it possible to quickly bring the highest quality services and applications to market at a dramatically lower cost. However, realizing the promise of NFV will require operators to transform both their networks and their business processes. DevOps principles are at the core of the business transformation, but DevOps needs to be extended to recognize the unique challenges of bridging development and operations processes in network operators. Operators’ gradual deployment of NFV further complicates these transformational challenges due to the need to test and assure both virtual and physical networks with a common approach.
Lifecycle Service Assurance builds on DevOps principles to enable operators to automate and unify testing across development and operations in physical, virtual, and hybrid networks. Operators who achieve the benefits of NFV in the shortest amount of time will be the ones who understand and aggressively address the unique challenges of transforming both their networks and their business.
Jeff Atkins is Director of Marketing at Spirent Communications, a multinational telecommunications testing company.
Filed Under: Infrastructure