Multi-Chassis Inter-Process Communication via Virtual Fabric Link
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data networks face challenges in achieving rapid convergence and resiliency due to high convergence times in Ethernet networks, particularly in edge and aggregate layers, which can lead to service disruptions and inefficiencies in traffic routing.
Innovation Solution
The implementation of a multi-chassis virtual-fabric link aggregation system that enables active/active forwarding, allowing all links to be actively forwarding traffic simultaneously, and facilitates inter-process communication between Aggregation Switches to manage traffic and control information efficiently, thereby reducing convergence times and enhancing network resiliency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional Ethernet networks use spanning tree protocol for redundancy, then network resiliency is provided, but convergence time becomes slow (several seconds)
Solution Approach 1:
The patent implements dynamic active/active forwarding where links transition between active and standby states based on real-time failure detection. The system dynamically switches from passive standby links to active forwarding links when failures occur, enabling rapid convergence without the slow spanning tree protocol reconfiguration process
Solution Approach 2:
The patent pre-configures multiple active links and standby links in advance, so that when a failure occurs, the standby links are already prepared for immediate activation. This preliminary preparation eliminates the need for spanning tree protocol to reconstruct paths from scratch, significantly reducing convergence time while maintaining network resiliency
2Device complexity
If only one path is active at a time in Ethernet networks, then protocol simplicity is maintained, but bandwidth efficiency decreases during failures
Solution Approach 1:
The patent maintains continuous useful action by keeping multiple links in active forwarding state simultaneously rather than switching between single active paths. This allows the network to continuously utilize available bandwidth on multiple paths, improving bandwidth efficiency during normal operation and during failures without significantly increasing protocol complexity
3Reliability
If redundant nodes and paths are provided for network resiliency, then availability increases, but device complexity increases
Solution Approach 1:
The patent segments the network into distinct active forwarding paths and standby backup paths, allowing independent management and simplification of each segment. This segmentation enables the system to maintain network availability through redundancy while reducing overall topological complexity by organizing redundant elements into structured, manageable groups rather than fully interconnected redundant topologies
Data Source
AI summary
Aggregation Switches connected via a virtual fabric link (VFL) are each active and able to communicate via an inter-process communication (IPC) using an IPC Virtual Local Area Network (IPC VLAN). A sub-set of Customer-facing ports (CFPs) of each Aggregation Switch are coupled to the VFL and assigned to the IPC VLAN to enable communication therebetween.