You have not logged in or are not authorized£¡
The networks and services of today demand greater bandwidth and enhanced processing capability from the control plane. Legacy network devices that use single frames and process routes in a centralized manner are lagging behind, as core routers are gradually evolving towards multi-frame, multi-control board, multi-CPU, and multi-core architecture.
As hardware has developed, network operating systems have evolved from single processing or multiprocessing into fully-distributed & fine-grained multiprocessing. Multi-CPU and multi-core hardware improves control plane processing, while ensuring reliability and scalability.
Huawei’s Versatile Routing Platform (VRP) is a network-level operating system that innovates and adapts to network development. It is a software platform, based on fully distributed architecture, that ensures performance, scalability, reliability and carrier-class management & maintenance.
Traditionally, every protocol or service is assigned a processing instance; this is restricted by the processing capability and memory space of a single CPU or core, and is now proving incapable of meeting growing demand.
Based on distributed architecture and industry-first RaaS (Routing as a Service) technology, Huawei VRP integrates various storage resources and dynamically balances the route calculation & storage loads. It also fully leverages both multi-core and multi-CPU parallel processing, so that system performance is greatly enhanced while O&M costs are reduced.
Figure 1 VRP RaaS
Huawei VRP enables smooth network upgrade. IP networking generally employs a variety of line cards that require upgrade to better interact with evolving hardware & chipsets, but large-scale software upgrade must be avoided when hardware or line cards are upgraded; software should also be forward and backwards compatible. Using an abstract driver frame and software-hardware decoupling technology, Huawei VRP ensures both for each line card.
Equipment reliability is a key issue. Huawei VRP uses Non-Stop x (NSx) and In-Service Software Upgrade (ISSU) technologies to provide comprehensive protection for network equipment; this greatly improves system reliability while ensuring carrier-class operations.
VRP supports Non-Stop Service (NSS), Non-Stop Routing (NSR), Non-Stop Bridge (NSB), and Non-Stop Forwarding (NSF), and delivers board-level NSx and process-level NSx so that full service-level NSx is implemented.
VRP also supports process-level ISSU; operators can select node-level (or board-level) ISSU or process-level ISSU as needed during deployment.
The VRP management plane uses a hierarchical architecture, where configuration tools, configuration models, and configuration data are separate. Based on the configuration model, the plane supports flexible expansion and a variety of configuration tools. The following figure illustrates the management plane architecture.
Figure 2 VRP management plane architecture
The configuration interface provides a variety of tools. User configuration requests are resolved using the configuration tools and then processed by the CFG component, which checks & associates them while generating configuration data according to the predefined model. After configuration is complete, the configuration data is stored in a central configuration database and can be obtained by the ARP database for each process.
This architecture greatly facilitates the deployment and maintenance of services. As an open, carrier-class, and multi-service network platform, Huawei VRP lays a solid foundation for the next generation of network operating systems.