Overview
As enterprises undergo digital transformation, data centers have increasingly complex services, while their scale and networking complexity also continue to grow. Amid this trend, enterprises are facing a host of challenges, such as low efficiency caused by manual deployment of multi-cloud networks, siloed management of multi-vendor devices, and complex configuration changes spanning multiple networks. iMaster NCE-Fabric—the core component of Huawei's L3.5 data center autonomous driving network—provides end-to-end (E2E) network services to quickly respond to requirements such as plug-and-play service provisioning and fast fault recovery. Specifically, it seamlessly interconnects networks with service workflows through the runbook in the northbound direction, and quickly adapts to multi-vendor devices through the Agile Open Container (AOC) in the southbound direction. This enables E2E automated network rollout with only minimal manual intervention.
Key Capabilities of NaaS

NaaS

NaaS provides southbound and northbound open programmability capabilities to support flexible service orchestration as well as unified management and control of multi-vendor devices. In this way, developers can quickly complete network orchestration to adapt to service changes, even without professional coding capabilities.

Key Features

  • Runbook

  • AOC

The runbook provides service-oriented network service integration. It supports visualized workflow orchestration based on RESTful APIs, and provides capabilities such as dry run, simulation, and full-process rollback, implementing service rollout within minutes.

Learn More

The AOC provides E2E southbound open service capabilities. Based on the YANG model, it supports device driver programmability, as well as unified deployment and simulation of hardware SDN, software SDN, traditional networks, and multi-cloud and multi-vendor networks, supporting zero-breakpoint orchestration of complex services.

Learn More
Benefits of NaaS
Flexible and Visualized Orchestration
Provides unified management, control, and visualization of heterogeneous multi-cloud networks, allowing users to orchestrate network resource packages and service flows through drag-and-drop operations.
Intelligent and Reliable Change
Performs dry run and simulation on the entire network to ensure zero human errors in change operations, and provides full-process rollback of service flows to quickly and accurately restore network configurations.
Open Programmability
Supports user-defined network resource packages and service flows, generates northbound RESTful APIs, and builds device drivers based on the generic NE model in the southbound direction to quickly adapt to multi-vendor devices.
Application Scenarios of NaaS
  • Day 0: Planning and Design
  • Day 1: Service Provisioning
  • Day N: Maintenance and Monitoring

Day 0: Planning and Design

lanning and Design Provides the open configuration capability before overlay service provisioning to implement fast network deployment and construction

iMaster NCE-Fabric provides the runbook capability to design and verify Day 0 work orders in advance. It then publishes them as northbound RESTful APIs so that third-party systems can complete Day 0 services such as switch capacity expansion, server capacity expansion, and server onboarding.

Day 1: Service Provisioning

Provides open atomic interfaces for service provisioning so that third-party systems can implement dynamic provisioning based on network resources.

iMaster NCE-Fabric provides the runbook capability to design and verify Day 1 work orders in advance. It then publishes them as northbound RESTful APIs so that third-party systems can complete Day 1 services such as route provisioning, policy provisioning, and VPC configuration.

Day N: Maintenance and Monitoring

Provides open atomic interfaces for maintenance and monitoring so that third-party systems can ensure stable running after function rollout.

iMaster NCE-Fabric provides the runbook capability to design and verify Day N work orders in advance. It then publishes them as northbound RESTful APIs so that third-party systems can complete Day N services such as device inspection and fault rectification

Development Process
Requirement Analysis
Perform requirement analysis and HLD based on service scenarios, and determine the service configuration logic and configurations to be delivered.
Action Model Development
Resource Package Development
(Optional) Develop a resource package based on the service configuration logic and configurations to be delivered.
Workflow Development
Runbook Creation
Create a runbook workflow.
Runbook Orchestration
Design and orchestrate the runbook workflow based on the service orchestration logic.
Runbook Publishing
Publish the runbook workflow. Only published workflows can be used to provision configurations.
Configuration Provisioning
Use the runbook workflow to create an instance and deliver configurations to devices.