Model-driven Management

Find out why model-driven interfaces are key enablers to simplify and automate network management

The massive increase in demand for 5G and cloud services is pushing the limits of network scaling, requiring network operators to manage more devices than ever before. In response, operators are beginning to adopt a NetOps strategy to simplify and automate network management.

Network automation and programmability are critical elements of NetOps that help operators increase operational efficiency and reduce costs. This shift to NetOps has been accompanied by the adoption of model-driven approaches to manage network elements, which have become key enablers for automation and programmability.

What is NetOps?

NetOps is a network operations strategy that brings automation to the network. It focuses on people, processes, and tools for deploying network services and applications to benefit both operators and end users. With a simpler, more reliable, and more efficient framework, optimal network performance and management is more easily achieved. Operators can deliver scalable and flexible networks and services in a timely and agile manner, while also providing an excellent quality of experience for customers.

What is model-driven management?

Model-driven management uses data models to define the structure, configuration, and behavior of various network elements. It lets operators add new network elements by integrating corresponding resource models without modifying management software, allowing network elements to become hot pluggable. Model-driven management decreases integration efforts across multiple vendors, products, and releases, reducing the cost of automating the management of network equipment. It also enhances network monitoring responsiveness for a faster mean time to recovery (MTTR) and faster time to revenue.

On this page

Overview
What is NetOps?
What is model-driven management?
What are YANG models?
Driving the move to a model-driven approach
What are the benefits of model-driven management?
Why deploy model-driven management now?
NetOps teams embrace model-driven approach
Related solutions and products
Learn more

What are Yet Another Next Generation (YANG) models?

YANG is a widely used data modeling language that enables operators to create consistent, rule-based behavior that is easy to adapt and integrate. Data models created with YANG make it possible for network devices to be configured using programmatic interfaces that use protocols such as the Network Configuration Protocol (NETCONF), gRPC Remote Procedure Call (gRPC), or the model-driven CLI (MD-CLI).

Network automation based on YANG models

YANG configuration: Programmability allows humans and machines to use programmatic interfaces to access structured data based on YANG models to configure network elements
YANG state: Visibility provides statistics and state information and includes the use of telemetry to push operational data to northbound management systems
YANG operations: Operations provides the ability to consistently manage network devices in automated network environments using programmatic interfaces

What is driving the move to a model-driven approach?

Operators have traditionally needed to create complex adapters that could translate between their management systems and vendor-specific command line interfaces (CLIs). This approach is cumbersome, error prone, and time-consuming. Operators had to invest significant time and effort to adapt to software releases for new equipment because CLIs often behave inconsistently and can easily break previous integrations. Model-driven network management addresses these shortcomings by simplifying network operations, improving efficiency, and enabling greater agility in deploying and managing network services.

What are the benefits of model-driven management?

Model-driven management provides operators with numerous benefits:

Enables operators to introduce new features and network elements quickly and easily into networks that are more dynamic and programmable

Consistent rule-based behavior reduces manual intervention and increases network resiliency

Standardized YANG-based programmatic interfaces ensure the same look and feel across all interfaces for easier network transformation and faster innovation

Standardized YANG-based data modeling language reduces the cost of automating the management of network equipment and the effort required to integrate hardware

Aligns the management interface to device functionality with a common human- and machine-readable configuration syntax, which reduces errors and improves operational efficiency

Leverages streaming telemetry, where network changes occur quickly and frequently, to optimize automation

Provides seamless mediation into a single multi-vendor, multi-technology automation platform, essential for network automation on a massive scale

Allows the functionality of the management protocol to access the device’s native functionality

Why deploy model-driven management now?

Model-driven approaches to network management have been in development for more than a decade. The technology is mature and provides a range of specialized tools to help operators transition smoothly and cost-effectively. Many operators recognize the significant ROI that comes with investing in model-driven solutions, along with the direct impact this approach has on users’ quality of experience. The programmability and automation capabilities enabled by model-driven network management make for faster innovation and easier network transformation, a key factor that gives operators a significant advantage over the competition and helps meet the goals of a NetOps strategy.

How do NetOps teams embrace the model-driven approach?

Nokia provides a variety of flexible solutions that enable operators to implement model-driven management approaches for IP routing and data center networks.

IP routing

Both IP routing operating systems, Nokia SR OS and SR Linux, include industry-leading model-driven management capabilities. The model-driven management of SR OS and the open, modular architecture of SR Linux provide configuration and state management through YANG-based MD-CLI, NETCONF, and gRPC and gNMI interfaces. They enable streaming telemetry through gRPC and gNMI subscriptions. They support operations through a YANG-based MD-CLI and NETCONF and support the Nokia Network Services Platform (NSP) through the northbound, model-driven REST interface.

The NSP uses a model-driven mediation software development kit and supports protocols such as NETCONF and gNMI in combination with YANG-defined data models to enable Nokia and multivendor equipment integration end-to-end across the network, to automate operations to reduce time to market and operational costs.

Data center networks

Nokia SR Linux is a truly open network operating system (NOS) that makes data center switching infrastructure more scalable, more flexible, and simpler to operate. SR Linux opens the NOS infrastructure with a unique architecture built using model-driven management and modern interfaces. This approach simplifies operations and integrations while giving operators the ultimate visibility and flexibility across their networks.

The Nokia Fabric Services System also enables the use of NetOps principles to simplify data center network operations. By using intent-based approaches, operators can program and implement network automation tasks at scale for each phase of their data center fabric operations, from Day 0 design to Day 1 deployment and Day 2+ operations.

The Nokia Network Developer Portal

The Nokia Developer Portal is designed to help developers transform operational practices and integrate network devices into their own infrastructure. The portal provides tips, examples, tutorials and labs to help developers use Nokia developed tools and third-party open-source tools to integrate with the Nokia products that run many of today’s global networks.

Visit the Nokia SR OS Developer Portal