Multi-Cluster & Fleet Management
In Development (Q4 2026)
Multi-cluster fleet management is currently in development, targeted for Q4 2026. This documentation describes the planned architecture and capabilities. Features are subject to change before release.
Purpose: For platform engineers and architects, explains what openCenter multi-cluster management provides — a single control point for visibility, policy, and lifecycle across your entire fleet.
The Problem
As organizations grow beyond a single cluster, operational complexity increases non-linearly:
- Visibility: Which clusters are healthy? Which are drifting?
- Policy: How do you enforce security baselines across 50 clusters?
- Lifecycle: How do you upgrade Kubernetes across a fleet without downtime?
- Consistency: How do you ensure every cluster has the same observability stack?
What Fleet Management Provides
| Capability | Description |
|---|---|
| Centralized visibility | Single dashboard showing health, versions, and drift status across all clusters |
| Policy distribution | Push Kyverno policies, RBAC templates, and security baselines fleet-wide |
| GitOps at scale | Manage fleet-level Kustomizations with cluster selectors and group targeting |
| Coordinated upgrades | Rolling Kubernetes and service upgrades with canary promotion |
| Cluster registration | Self-service cluster enrollment with token-based authentication |
| Federated observability | Aggregated metrics, logs, and alerts across the fleet |
Architecture
openCenter fleet management uses a hub-spoke model:
Hub Cluster (management plane)
├── Fleet Controller
├── Policy Distributor
├── Federated Prometheus
└── Fleet Dashboard
Spoke Cluster A ──agent──► Hub
Spoke Cluster B ──agent──► Hub
Spoke Cluster C ──agent──► Hub
Related Pages
- Architecture — hub-spoke design details
- Registration — enrolling clusters
- GitOps — distributing configurations
- Policies — centralized enforcement
- Observability — federated monitoring
- Upgrades — coordinated rollouts