Root Cause Analysis Method Based on Large Language Models with Residual Connection Structures
This addresses root cause analysis for operators in complex microservice systems, representing an incremental improvement over existing methods.
The paper tackles root cause localization in microservice architectures by proposing RC-LLM, a method using large language models with residual connections to integrate multi-source data and model causal dependencies, achieving strong accuracy and efficiency on CCF-AIOps datasets.
Root cause localization remain challenging in complex and large-scale microservice architectures. The complex fault propagation among microservices and the high dimensionality of telemetry data, including metrics, logs, and traces, limit the effectiveness of existing root cause analysis (RCA) methods. In this paper, a residual-connection-based RCA method using large language model (LLM), named RC-LLM, is proposed. A residual-like hierarchical fusion structure is designed to integrate multi-source telemetry data, while the contextual reasoning capability of large language models is leveraged to model temporal and cross-microservice causal dependencies. Experimental results on CCF-AIOps microservice datasets demonstrate that RC-LLM achieves strong accuracy and efficiency in root cause analysis.