Intelligent and Highly Available IP Design Concepts – Design of the Product System Architecture
Ever since the introduction of the first 10GE switch to the market in 2003, Ruijie Networks Co., Ltd. has been continuously dedicated to the construction of stable core network solutions. After six years of experiences in the market when Ruijie products have been deployed by over 6000 customers, Ruijie Networks has gradually developed the intelligent and highly available IP design concepts over the procedures of product development, system architecture, and technical development.
Based on a stable hardware architecture and modular software designs, the intelligent and highly available IP design concepts of high-end switching ensure high reliability in various networking models through intelligent protection technologies of network topologies. In addition, the intelligent network monitoring and maintenance technologies of Ruijie high-end switching effectively address network faults and decrease the network down times, transmitting user services in a steady manner.
Design of the Product System Architecture
In the presence of frequent blue screens and crashes, antivirus software alone does not ensure the steady operation of operating systems. It makes practical sense only when stability is fully considered in the design of the product system architecture and highly available network technologies are applied.
Product System Architecture – Design of the Stable Hardware Architecture
In addition to carrying software, hardware is also an indispensable part in data switching, protocol processing, and device management. Hardware instability can directly result in the disaster of the device as a whole.
Stability is taken into full account in the design of the Ruijie 10GE switching hardware architecture.
Product System Architecture – Edges of the Ruijie Hardware Architecture Design
1.        The data forwarding plane is strictly separated from the network protocol control plane and different service data (such as management data, protocol packets, and forwarded packets) are processed by different planes, thereby ensuring a stable and reliable lower layer.
The line card is connected to the engine CPU and CPLD through internal communication modules, forming the control plane responsible for device management, information synchronization, and network protocol processing.
The Crossbar chip and the ASIC chip on the line card form the data plane. The line card ASIC is responsible for handling service data whereas the Crossbar takes charge of high-speed internal forwarding.
2.       Distributed service processing: The functions of device management and service processing are distributed to each line card rather than integrated on the processing engine, thereby reducing the engine pressure and meanwhile ensuring stability.
The distributed CPU, CPLD design, the line card and the engine globally synchronize the Layer 3 routing, Layer 2 forwarding or device management information to bring high reliability of the control plane.
In distributed service processing, each line card (ASIC chip) supports IPv4/IPv6/ACL/QoS services. On the basis of global synchronization, the Crossbar is connected to the line card ASIC to forward internal data at high speed, ensuring high reliability on the data forwarding plane.
Product System Architecture