Zhang Jingru

Inner Mongolia Agricultural University, Vocational and Technical College, Network Engineering Major

Abstract：

As cloud-native deployment extends across regions, cloud platforms, hybrid infrastructures, and edge environments, Kubernetes multi-cluster management has become a key approach for enterprise infrastructure evolution. The shift from single-cluster orchestration to cross-cluster governance raises new requirements for control-plane responsibility division, state coordination, network interconnection, security governance, and fault recovery. The review develops a classification framework, mechanism comparison, and scenario adaptation recommendations for Kubernetes multi-cluster control plane management. Multi-cluster control plane architectures are classified into centralized control plane, distributed control plane, and cloud-edge master-slave collaboration, and their mechanisms are compared in terms of state synchronization, joint scheduling, cross-cluster networking, security governance, heterogeneous adaptation, and fault-tolerant recovery. No universal optimal architecture exists; architecture selection should be aligned with business scale, deployment location, network conditions, resource heterogeneity, and operation and maintenance capability.

Key Words：

Kubernetes; multi-cluster; control plane management; cloud native; architecture optimization