Journal of Intelligent Computing Kai Wang, Na Wang 15 4 2024 https://doi.org/10.6025/jic/2024/15/4/135-141 https://www.dline.info/jic/fulltext/v15n4/jicv15n4_3.pdf The rapid progress of cloud computing technology has attracted widespread attention, and the importance of microservices architecture is also increasing. In this regard, we studied a new algorithm, Xgboost, which effectively addresses the load balancing issues in the current microservices cluster, thus better meeting user demands. We identified factors that significantly impact load balancing effectiveness through in-depth research on various features. We used ensemble learning to estimate the load received by each server node, thereby achieving load balancing. Experimental results confirm that our proposed algorithm significantly improves throughput, reduces interception errors, and lowers the system’s average response time compared to other load-balancing algorithms.