Ethernet Ring Protection Switching via Selective FDB Flushing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In Ethernet ring networks, frequent flushing of Filtering Databases (FDBs) due to link failures leads to traffic overshoot as all nodes need to relearn MAC source addresses, causing inefficiency and increased traffic.
Innovation Solution
A method for selective FDB flushing based on whether the failed link affects data forwarding, using a Do Not Flush (DNF) bit in protection switching messages to determine which nodes should flush their FDBs, thereby minimizing unnecessary FDB updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all nodes flush their FDBs upon receiving SF messages, then forwarding tables are updated to reflect network changes, but traffic overshoot increases due to frequent MAC source address learning
Solution Approach 1:
The patent applies local quality by making FDB flushing selective rather than universal. Nodes determine whether to flush their FDBs based on whether they are affected by the link failure, with the DNF bit in SF messages indicating which nodes should preserve their forwarding tables. This resolves the contradiction by updating only the necessary local FDBs while leaving others intact, maintaining forwarding accuracy where needed while avoiding unnecessary traffic overshoot.
2Adaptability or versatility
If FDBs are flushed frequently to maintain accurate forwarding tables, then network adaptability improves, but network efficiency deteriorates due to increased traffic congestion
Solution Approach 1:
The patent implements local quality by differentiating which nodes need to adapt their FDBs based on their relationship to the failed link. The DNF bit mechanism allows the network to maintain adaptability in affected areas while preserving efficiency in unaffected areas, avoiding blanket FDB flushing that would degrade overall network productivity.
3Measurement precision
If all nodes perform MAC source address learning after FDB flush, then forwarding path accuracy is restored, but network traffic increases due to relearning overhead
Solution Approach 1:
The patent extracts the FDB flushing action from being a universal response and makes it selective. By using the DNF bit to identify which nodes should flush their FDBs, the system removes the unnecessary FDB flushing and associated MAC learning traffic from nodes that are not affected by the link failure, thereby reducing overall network traffic while maintaining forwarding accuracy where it matters.
Data Source
AI summary
A method for protection switching in an Ethernet ring network is provided. According to an aspect, the protection switching method includes: at a first node on the Ethernet ring network, detecting link failure; at the first node, determining whether a port through which a protection switching message has to be transmitted is on a forwarding path; and at the first node, generating a protection switching message based on the result of the determination; and at the first node, transmitting the protection switching message.