Stacking System Traffic Switching Topology Change

Solution Overview

Problem

Current stacking systems experience traffic interruption when transitioning from a daisy chain to a ring topology due to the need for devices to recalculate their shortest forwarding paths, which can lead to inconsistent and incomplete traffic switching.

Innovation Solution

The method involves blocking a stack port of newly joined devices to maintain the daisy chain topology until their configurations are restored, then calculating and sending notification packets for topology changes from daisy chain to ring, ensuring all devices recalculate their paths simultaneously and switch traffic without interruption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If the master restores configuration of newly joined devices and triggers topology change notification immediately, then the ring topology is established quickly, but traffic interruption occurs because devices recalculate shortest forwarding paths asynchronously and incompletely

Engineering Contradiction:
Core Design Contradiction:
SpeedVSReliability

Solution Approach 1:

The patent applies preliminary action by blocking the stack port of the newly joined device before restoring its configuration. This preliminary blocking ensures that the topology remains in a stable daisy chain state during configuration restoration, preventing premature traffic switching that would cause interruption. The port is blocked in advance, and only opened after configuration restoration is complete, ensuring consistent topology state across all devices.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the topology convergence process into distinct phases: first completing the daisy chain topology convergence with the newly joined device, then restoring configuration, and finally triggering ring topology convergence. This segmentation separates the operations in time, ensuring that each phase completes fully before the next begins, thereby preventing traffic interruption during transitions.

Inventive Principle:
Principle #1Segmentation

2Productivity

If devices recalculate shortest forwarding paths independently without coordination, then each device optimizes its own routing, but inconsistent forwarding paths cause traffic interruption during topology transition

Engineering Contradiction:
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent uses feedback mechanisms through topology change notification packets that propagate through the network. When one device detects a topology change and recalculates its forwarding path, it sends notification packets to other devices, creating a feedback loop that coordinates recalculation across the entire network. This ensures all devices update their forwarding paths consistently rather than independently.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent merges the topology change detection and notification functions into a coordinated process. The master device combines configuration restoration with topology change detection, and the blocked stack port mechanism merges the topology maintenance function with the configuration restoration process. This integration ensures that all devices operate with consistent topology information during transitions.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS8248969B2Stacking system and a method for switching traffic in the stacking system
Publication Date: 2012.08.21
  • US8248969B2 patent drawing
  • US8248969B2 patent drawing
  • US8248969B2 patent drawing

AI summary

The invention provides a stacking system and a method for traffic switching in the stacking system, comprising: when newly joined devices in the stacking system make the daisy chain topology change to ring topology, block a stack port of one of the newly joined devices; after the devices in the stacking system perform another topology convergence process, the master restores the configuration of the newly joined devices; after the configuration of the newly joined device one of whose stack ports is blocked is restored, it sets its blocked stack port to the forwarding state, calculates the shortest path based on the ring topology, and sends first notification packets for topology change from daisy chain type to ring type through its two stack ports; each device that receives the first notification packets re-calculates its shortest forwarding path respectively and switches the traffic, and then forwards the first notification packets. This invention can prevent traffic interruption that occurs when the topology of a stacking system changes from daisy chain to ring.