Virtual Chassis Network Resiliency via Unified Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data networks face challenges in maintaining high availability and network resiliency due to difficulties in upgrading or transitioning between different node architectures and topologies, which affects the ability to handle component and link failures effectively.
Innovation Solution
A virtual chassis system is introduced, where network nodes operate as a single logical device with unified management, using virtual fabric links for communication and synchronization, allowing for seamless transition between different topologies and architectures, and enabling active forwarding of traffic across multiple links for increased redundancy and bandwidth efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If network nodes use different node architectures and topologies to provide redundancy, then network resiliency is improved, but the difficulty of upgrading or transitioning between architectures increases
Solution Approach 1:
The system segments network nodes into different architectural types (single switching element, stackable switching elements, multi-slot chassis) while providing a unified control plane that manages each type independently. This segmentation allows diverse architectures to coexist and be upgraded independently without affecting the entire network, resolving the contradiction between providing resiliency through architectural diversity and maintaining ease of upgrade.
Solution Approach 2:
A virtual chassis controller acts as an intermediary between network nodes with different architectures. It provides a unified control interface that abstracts the underlying architectural differences, enabling seamless transition and migration between different node types without requiring direct compatibility between all node architectures, thus reducing upgrade difficulty while maintaining resiliency.
2Reliability
If multiple physical paths are provided between nodes to prevent single points of failure, then network resiliency is improved, but the complexity of managing topology transitions increases
Solution Approach 1:
The control plane is designed with universal functionality to manage multiple topology types (ring networks, partial mesh networks, full mesh networks, hub networks) and multiple node architectures through a single unified interface. This multi-functionality eliminates the need for separate management mechanisms for each topology type, reducing management complexity while maintaining comprehensive resiliency capabilities.
Solution Approach 2:
The system manages topology transitions by dynamically changing control parameters such as active paths, convergence thresholds, and failure detection intervals rather than requiring complex reconfiguration procedures. This parameter-based approach simplifies topology management by transforming structural changes into parameter adjustments, reducing the complexity of managing multiple physical paths.
3Reliability
If convergence time is reduced to less than one second for seamless transition, then network availability is improved, but the detection and convergence mechanisms become more complex
Solution Approach 1:
The system performs preliminary actions by pre-configuring multiple alternate paths and pre-computing convergence procedures before failures occur. When a failure is detected, the pre-configured paths are immediately activated without requiring complex real-time computation, achieving sub-second convergence times while simplifying the actual failure response mechanism through advance preparation.
Solution Approach 2:
The control plane implements feedback mechanisms that continuously monitor network status and automatically trigger convergence procedures when threshold violations occur. This automated feedback loop eliminates manual intervention and complex decision-making during failure events, reducing convergence time by eliminating delays associated with detection and manual configuration while keeping the mechanisms manageable through systematic monitoring.
Data Source
AI summary
A virtual chassis system includes a plurality of network nodes configured with a master virtual chassis address. The network nodes are connected by virtual fabric link (VFLs) that provide a connection for exchange of data packets between the network nodes. The data packets include source MAC addresses and associated hardware device information, such as source chassis ID, source network interface identifier and source port identifier information. The network nodes use this information to maintain synchronized MAC address tables for forwarding of data packets in the virtual chassis system. One or more control protocols in the network node are used for topology discovery, master network node election, generation of routing tables, health monitoring and other functions.