#Overview

A cluster is the foundational resource collection for running containerized applications, encompassing nodes, load balancers, storage, and other critical components. It is a prerequisite for successfully running containerized applications on the platform. During initial platform installation, a standard Kubernetes cluster, known as the global cluster, is created. Subsequently, multiple clusters can be integrated into the global cluster for unified management.

#Cluster Types

#On-Premises Clusters

On-Premises clusters are Kubernetes clusters directly created by the platform. Users provide virtual or physical machines, and the platform installs and configures Kubernetes clusters on these machines. This approach is suitable for enterprises with existing hardware resources, allowing full utilization of infrastructure.

#Managed Clusters

Managed clusters are Kubernetes clusters provided by cloud service providers, which are integrated into the platform for unified management. Supported integration methods include:

Method	Description	Use Case	Key Characteristics
Import	Integrating existing Kubernetes clusters	Existing clusters with direct network access	Cluster information submitted to global cluster global cluster must have network access to the cluster
Register	Integrating clusters with strict security requirements	Clusters with high security constraints	Specific plugins installed on the target cluster Reverse proxy establishes a secure tunnel Maintains cluster security while enabling management
Proxy Create	Creating clusters through cloud service providers	Leveraging public cloud Kubernetes services	Cloud service provider credentials required Platform creates Kubernetes clusters using provided credentials

#Multi-Cloud and Hybrid Cloud Support

These cluster management approaches meet enterprise needs in multi-cloud and hybrid cloud scenarios, supporting container transformation at different stages:

• Existing Hardware: Create platform-provided clusters
• Existing Clusters: Import or register into the platform
• Elastic Demands: Quickly create public cloud clusters

#Implementation Considerations and Limitations

#Version Compatibility

• Supported Kubernetes versions: 1.28, 1.29, 1.30, 1.31
• Both On-Premises and Managed clusters must ensure version compatibility
• Version mismatches may result in feature limitations or compatibility issues

#Network and Security Requirements

• Ensure network connectivity between global and target clusters
• Implement appropriate firewall and network security policies
• Manage access credentials and authentication mechanisms securely

#Best Practices for Cluster Management

#1. Pre-Implementation Assessment

• Conduct thorough infrastructure and workload analysis
• Identify specific requirements for each cluster
• Develop a comprehensive migration and integration strategy

#2. Security and Compliance

• Implement role-based access control (RBAC)
• Use network policies to restrict cluster communication
• Regularly audit and update security configurations
• Ensure compliance with industry standards and regulations

#3. Monitoring and Observability

• Set up centralized logging and monitoring
• Implement proactive alerting mechanisms
• Use platform-provided observability tools
• Track cluster performance, resource utilization, and health

#4. Backup and Disaster Recovery

• Establish regular backup procedures
• Create and test disaster recovery plans
• Implement multi-cluster backup strategies
• Ensure minimal downtime and data loss

#5. Continuous Optimization

• Regularly review cluster configurations
• Optimize resource allocation
• Update to the latest supported Kubernetes versions
• Leverage platform features for automatic updates and scaling