Microservice root cause positioning method and system based on fault snapshot and large language model
By combining fault snapshots with a large language model, the problem of rapid and accurate fault location in microservice architecture is solved, achieving efficient fault repair and system stability improvement, and providing a global diagnostic foundation and deep causal reasoning capabilities.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NARI TECH CO LTD
- Filing Date
- 2026-01-20
- Publication Date
- 2026-06-05
AI Technical Summary
In complex microservice architectures, existing technologies struggle to quickly and accurately pinpoint the root cause of failures, especially cross-domain cascading failures. This results in long repair times, complex operations and maintenance, and reliance on expert experience. Traditional methods lack causal reasoning capabilities and data fusion mechanisms, making it impossible to effectively handle complex implicit causal relationships.
A microservice root cause localization method based on fault snapshots and large language models is adopted. The method captures abnormal alarms through a monitoring and alarm agent, aggregates multi-source telemetry data, executes an aggressive preprocessing strategy to generate structured fault snapshots, and uses a large language model trained with a specific composite to perform deep causal reasoning to generate a structured root cause analysis report.
It achieves rapid and accurate fault location, reduces fault repair time, improves system stability and reliability, with an accuracy rate of over 95%, reduces operation and maintenance costs, and can automatically handle root cause location of complex call chains.
Smart Images

Figure CN122152565A_ABST