GAL-MAD: Towards Explainable Anomaly Detection in Microservice Applications Using Graph Attention Networks

The distributed and dynamic nature of microservices poses significant challenges to maintaining system reliability, highlighting the need for effective anomaly detection. Existing statistical and classical machine learning methods often fail to capture the high-dimensional dependencies and complex interactions found in these architectures. Public datasets are also limited, with few offering the multivariate performance metrics needed for realistic evaluation. This work introduces the RS-Anomic dataset generated using the open-source RobotShop microservice application. The dataset captures multivariate performance metrics and response times under normal conditions and anomalous conditions, encompassing ten types of anomalies. We propose a novel anomaly detection model called Graph Attention and LSTM-based Microservice Anomaly Detection (GAL-MAD), leveraging Graph Attention and Long Short-Term Memory architectures to capture spatial and temporal dependencies in microservices. Using SHAP, we explore anomaly localization to enhance explainability. Experimental results demonstrate that GAL-MAD outperforms state-of-the-art models on the RS-Anomic dataset, achieving higher accuracy and recall across varying anomaly rates.