While most industry experts agree that Network Functions Virtualization (NFV) technology is an important step in network evolution and transformation, many communications service providers (CSPs) are still wrestling with real-world issues and challenges related to commercial deployments.
Yet even in the early stages of the NFV growth curve, there are several key areas where NFV will add business value in 2017. This is the latest report from our NFV Playbook and it is entitled “Six Things that NFV is good for in 2017”. This exclusive report has been authored by Sue Rudd, Director Service Provider Analysis at Strategy Analytics. It looks at how NFV is starting to make a real difference in six important ways:
- NFV Application #1: Metro Area Software Defined Wide Area Network (SD-WAN)
SD-WAN for ‘on demand’ metro area data center connectivity or hybrid cloud traffic load management. This can drive widespread use of SDN/NFV for customer ‘control’ of enterprise WAN and VPN services to generate revenues. It also dramatically simplifies and reduces provisioning cost of enterprise VPNs.
- NFV Application #2: Virtual Customer Premise Equipment (vCPE)
Soft configurable, maintainable customer premise equipment (CPE) enables improved service flexibility and simple layer 2 connectivity for SMEs and consumers. It’s no wonder that CSPs are embracing NFV to manage customer termination equipment, saving equipment CAPEX with centralized software.
- NFV Application #3: Pre-NFV/SDN-based virtualized legacy infrastructure equipment
Standardized configuration management helps avert around 30% of network problems that result from configuration inconsistencies. Moreover, dynamic resource allocation can improve capacity utilization by up to 60%, as well as boosting services at peak periods to improve Quality of Experience (QoE).
- NFV Application #4: NFV telco data centers using SDN controllers
Network providers can change operations processes for VMs based on proven stable legacy appliances within ‘Cloud Ready’ NFV data center functionality. What’s more, VNF components for integrated standard IMS service logic chains and core voice services can be scaled for Virtual Machines as required.
- NFV Application #5: Focus on ‘operations first’ strategy
Many CSPs are first implementing standardized versions of new operations processes before new VNF deployments. This OSS first strategy helps eliminate over 65% of network problems with new NFV configuration and service management. Also, automated configuration of standard ‘mix and match’ components reduces configuration errors, improves network performance, and lowers OPEX.
- NFV Application #6: New agile processes for faster service creation and activation
Operators can achieve faster time to market for new services and customer service modifications with new agile processes leveraging NFV. VNFs can be implemented quickly as dynamic service chains for instant ‘on-demand bandwidth’ activation. CSPs also can configure new VNF service logic flows rapidly and store pre-tested new service features for variable combinations of voice and Unified Communications service VNFs.
Walk, Don’t Run
Before we get too carried away with dreams of NFV success, let’s remember to pace ourselves. CSPs who try to deploy brand new NFV functionality while simultaneously adopting new operations processes, can find themselves stalled in mid-air. In fact, deploying new processes and infrastructure changes at the same time is a little like changing the engine and flaps on a large aircraft mid-flight. Not a good idea, in other words!
When service providers try to change both VNF functional elements and the NFV management or operations processes simultaneously, the result can be an unstable network that is difficult to troubleshoot. Those operators who are moving fastest with NFV deployment are typically moving in parallel. They test the new processes that NFV requires and in parallel introduce new VNF based service functions for a stable service, such as SD-WAN, vCPE and vEPC, before bringing the new processes and the new software together.
Unlike popular DevOps processes, which proceed in a continuous integration fashion, step function NFV implementation runs on two parallel paths. The paths are then converged and the process is repeated. One path is driven by operations to achieve immediate cost savings – and skills enhancement – with the implementation of new processes. The second, parallel path is driven by network and service deployment of a single proven VNF based service, such as SD-WAN or vCPE, requiring a limited set of process changes to support a controlled service rollout with limited scope. In this way, limited regional service can be deployed and then extended virtually to a wider area, one step at a time.
To learn more about how to avoid missteps while successfully deploying NFV technology for real business value, register for The NFV Playbook 2017 today.