MPLS Link Fault Detection via LSP Unavailability Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional link fault detection methods in MPLS networks fail to accurately determine the overall fault state of multiple links used for service data transmission, leading to data packet loss and reduced reliability.
Innovation Solution
A method where a network device sends data traffic over multiple label switched paths (LSPs), determines unavailability based on missed detection packets, and stops data transmission when unavailability exceeds a threshold, allowing for selection of a less faulty path for continued data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional BFD or OAM method is used for link detection, then each individual link fault can be detected, but the overall fault state of multiple LSPs for a service flow cannot be accurately determined
Solution Approach 1:
The patent combines the fault detection results of multiple individual LSPs into an overall unavailability assessment. By merging the detection outcomes of each LSP path and calculating a composite unavailability metric, the system achieves accurate determination of the overall fault state for service flow data transmission, resolving the contradiction between individual link detection capability and overall path reliability assessment.
2Reliability
If multiple LSPs are used to transmit service data, then data transmission reliability can be improved through path diversity, but the complexity of detecting and managing fault states increases
Solution Approach 1:
The patent transforms the complex multi-path fault detection problem into a parameter calculation problem. By changing the detection focus from individual link states to an aggregated unavailability parameter, and by defining clear calculation rules for LSP unavailability based on detection packet reception, the system simplifies the management of fault states across multiple paths while maintaining high reliability.
3Loss of time
If link fault detection is performed using conventional methods, then detection packets are sent periodically, but data packet loss occurs when multiple LSPs are faulty
Solution Approach 1:
The patent implements preliminary fault detection before data transmission failures occur. By continuously monitoring LSP availability through detection packets and calculating unavailability in advance, the system can take preventive actions to switch data paths before actual data packet loss happens, thus eliminating the contradiction between detection timing and data loss prevention.
Data Source
AI summary
A link fault detection method includes sending, by a first network device, data traffic to a second network device by using a plurality of first label switched paths (LSPs). The method also includes determining, by using each of the plurality of first LSPs within a preset period, whether a first link detection packet from the second network device is received. The method further includes determining an LSP of the plurality of first LSPs for which the first link detection packet is not received as a first faulty LSP. The method additionally includes calculating a first unavailability of the plurality of first LSPs based on the first faulty LSP. The method also includes triggering the plurality of first LSPs to stop transmitting the data traffic according to the first unavailability of the plurality of first LSPs.