Flow-Graph Based Optimization of Containerized Resource Scheduling: An Efficient Scheduling Algorithm for Kubernetes Clusters

Flow-Graph Based Optimization of Containerized Resource Scheduling: An Efficient Scheduling Algorithm for Kubernetes Clusters

연구 분야: Software Development

학회: 2024 International Conference on Intelligent Computing and Robotics (ICICR)

With the widespread adoption of containerization technology and the evolution of cloud computing, efficient scheduling of containerized applications in cluster environments has become an increasingly important area of research. Kubernetes, as a mainstream container orchestration tool, faces several challenges with its default queue-based and greedy scheduling algorithm, especially in large-scale deployments, such as high latency and low scheduling robustness. This paper introduces a flow-graph-based optimization method for containerized resource scheduling, modeling the scheduling problem as a Minimum Cost Maximum Flow problem to achieve global optimization. Comparative experiments with traditional queue-based and random scheduling algorithms demonstrate the superiority of the proposed method in terms of balanced resource utilization, robustness of scheduling outcomes, and reduced algorithmic latency. This study provides a more efficient resource scheduling strategy for container orchestration tools like Kubernetes and offers a new perspective on resource management in large-scale containerized application environments.