Optimized Intention-adaptive Graph Neural Network for Robust Failure Diagnosis of Microservice System Using Multimodal Data

Abstract Automatic failure diagnosis is critical for large-scale Microservice systems. Most failure detection methods available today only employ single-modal data (logs, traces, or metrics). It carries out an empirical study using actual failure scenarios to show that the integration of several data sources (multimodal data) leads to a more precise diagnosis. Effectively expressing these data and handling unequal failures remain challenging. The suggested paper, MD-RFD-MS-IGNN, proposes Multimodal Data for Robust Failure Diagnosis of Microservice System using Optimized Intention-adaptive Graph Neural Network. First, GAIA dataset is used to collect the input data. To implement this, the input data is pre-processed using Adaptive Kernel Learning Kalman Filtering (AKLKF) and it removes the N/A (not applicable) values or empty row from the collected data; then the Pre-processed data are feature extracted using Automated Tunable Q Wavelet Transform (ATQWT)is used to extract spatial features like strace, log, and metric. Subsequently, the gathered data is loaded into an Intention-adaptive Graph Neural Network (IGNN) to efficiently classify failure detection into categories such as file missing, system stuck, process crash, and access refused. Generally speaking, optimization algorithms that may be modified to get the optimal parameters for accurate failure classification are not expressed by IGNN. In order to optimize Intention-adaptive Graph Neural Networks, which precisely identify Microservice system failure, Human Evolutionary Optimization (HEO) was used. Next, the suggested MD-RFD-MS-IGNN is put into practice, and performance measures including F1-Score, Precision, and Recall are examined. When analysed through existing techniques such as Automated functional and robustness testing of microservice architectures (AFRT-MA-GCM), MTG_CD: multi-scale learnable transformation graph for fault classification and diagnosis in microservices (MTG-FCDM-GCN), and robust failure diagnosis of microservice system through multimodal data (RFD-MS-MD-GNN), the performance of the MD-RFD-MS-IGNN approach achieves 17.30%, 23.39%, and 32.41% higher F1-Score.