Skip to main content

Cluster Management

KubeVision can manage multiple Kubernetes clusters from a single dashboard. All clusters share the same UI, RBAC, and audit log.

Auto-Detection from kubeconfig​

On startup, KubeVision reads the local kubeconfig (~/.kube/config by default) and registers every context it finds as a managed cluster.

# Override the kubeconfig path via environment variable
KUBECONFIG=/path/to/kubeconfig ./kubevision
tip

If your kubeconfig has multiple contexts, all of them are imported automatically. You do not need to add them manually.

Adding a Cluster​

  1. Go to Settings → Clusters
  2. Click Add Cluster
  3. Fill in the form:
FieldDescription
NameDisplay name shown in the sidebar
Authentication TypeUse Kubeconfig for local or external clusters; use In-Cluster only when KubeVision runs inside the target cluster
kubeconfigPaste the kubeconfig YAML or upload the file
  1. Click Add Cluster — KubeVision validates connectivity before saving

For a local k3d cluster, export a standalone kubeconfig with k3d kubeconfig get <cluster-name> and paste or upload that output. Kubeconfig contents contain client credentials and should not be committed to source control.

Removing a Cluster​

  1. Go to Settings → Clusters
  2. Find the cluster row and click the Delete icon
  3. Confirm the dialog
warning

Removing a cluster deletes all cached resource data for that cluster. Audit logs and favorites referencing it are retained.

Switching Between Clusters​

Use the cluster selector in the top navigation bar to switch the active cluster. Resource lists, search results, and topology views are scoped to the selected cluster.

Cluster Health Status​

The sidebar shows a live health badge next to each cluster name:

BadgeMeaning
Green dotAPI Server reachable, informer connected
Yellow dotDegraded — partial connectivity or stale cache
Red dotUnreachable — all reads fall back to cached data

Health is checked every 30 seconds via a lightweight GET /healthz probe against the API Server.