Ciena is adding policy software from the Linux Foundation’s open-source Open Network Automation Platform (ONAP) to its Blue Planet network management platform.
Operators want to use automation to help tackle the growing complexity and cost of operating their networks.
Kevin Wade“Policy plays a key role in this goal by enabling the creation and administration of rules that automatically modify the network’s behaviour,” says Kevin Wade, senior director of solutions, Ciena’s Blue Planet.
Incorporating ONAP code to enhance Blue Planet’s policy engine also advances Ciena’s own vision of the adaptive network.
Automation platforms
ONAP and Ciena’s Blue Planet are examples of network automation platforms.
ONAP is an open software initiative created by merging a large portion of AT&T’s original Enhanced Control, Orchestration, Management and Policy (ECOMP) software developed to power its own software-defined network and the OPEN-Orchestrator (OPEN-O) project, set up by several companies including China Mobile, China Telecom and Huawei.
ONAP’s goal is to become the default automation platform for service providers as they move to a software-driven network using such technologies as network functions virtualisation (NFV) and software-defined networking (SDN).
Blue Planet is Ciena’s own open automation platform for SDN and NFV-based networks. The platform can be used to manage Ciena’s own platforms and has open interfaces to manage software-defined networks and third-party equipment.
Ciena gained the Blue Planet platform with the acquisition of Cyan in 2015. Since then Ciena has added two main elements.
One is the Manage, Control and Plan (MCP) component that oversees Ciena's own telecom equipment. Ciena’s Liquid Spectrum that adds intelligence to its optical layer is part of MCP.
The second platform component added is analytics software to collect and process telemetry data to detect trends and patterns in the network to enable optimisation.
“We have 20-plus [Blue Planet] customers primarily on the orchestration side,” says Wade. These include Windstream, Centurylink and Dark Fibre Africa of South Africa. Out of these 20 or so customers, one fifth do not use Ciena’s equipment in their networks. One such operator is Orange, another Blue Planet user Ciena has named.
A further five service providers are trialing an upgraded version of MCP, says Wade, while two operators are using Blue Planet’s analytics software.
In a closed-loop automation process, the policy subsystem guides the orchestration or the SDN controller, or both, to take actions
Policy
Ciena has been a member of the ONAP open source initiative for one year. By integrating ONAP’s policy components into Blue Planet, the platform will support more advanced closed-loop network automation use cases, enabling smarter adaptation.
“In a closed-loop automation process, the policy subsystem guides the orchestration or the SDN controller, or both, to take actions,” says Wade. Such actions include scaling capacity, restoring the network following failure, and automatic placement of a virtual network function to meet changing service requirements.
In return for using the code, Ciena will contribute bug fixes back to the open source venture and will continue the development of the policy engine.
The enhanced policy subsystem’s functionalities will be incorporated over several Blue Planet releases, with the first release being made available later this year. “Support for the ONAP virtual network function descriptors and packaging specifications are available now,” says Wade.
The adaptive network
Software control and automation, in which policy plays an important role, is one key component of Ciena's envisaged adaptive network.
A second component is network analytics and intelligence. Here, real-time data collected from the network is fed to intelligent systems to uncover the required network actions.
The final element needed for an adaptive network is a programmable infrastructure. This enables network tuning in response to changing demands.
What operators want, says Wade, is automation, guided by analytics and intent-based policies, to scale, configure, and optimise the network based on a continual reading to detect changing demands.