Algorithm for Locating Fault Linkage Sets in End-to-End Microservice Networks Under Multiple Business Scenarios

As microservice architectures become increasingly prevalent in enterprise IT systems, the need for an effective fault management and diagnostic framework is paramount. This paper introduces a sophisticated algorithm capable of detecting fault linkage sets within extensive microservice networks. Integrating graph theory with machine learning, specifically an enhanced random forest model, our approach adeptly constructs service dependency graphs, unveiling dynamic microservice connections. This facilitates the accurate prediction and localization of potential faults. Rigorous testing across diverse business scenarios has confirmed that our algorithm excels in fault detection accuracy and response time when benchmarked against existing methods. Our findings illustrate the algorithm's adeptness in adapting to changing business demands via real-time monitoring and dynamic adjustment. The synergy between latency, traffic, and error rates-as revealed through our correlation analysis-provides insights pivotal for proactive system maintenance. This research contributes a refined methodology to the domain of microservice fault localization, offering a robust tool for sustaining system stability and promoting efficient operational management.