Your first cluster

KubeDrive reads the same kubeconfig as kubectl: the files in $KUBECONFIG, or ~/.kube/config when it is not set. There is nothing to install on the cluster and nothing to import: every context appears on the Clusters screen, and edits made with kubectl show up without a restart.

The Clusters screen

Contexts are grouped into Cloud (EKS, GKE, AKS and other managed clusters), Other and Local (kind, minikube, Docker Desktop and the like). Each one shows its status:

• Reachable with the server’s Kubernetes version;
• Address not found or No response — usually a VPN you need to turn on;
• No AWS profile — the profile the context uses is missing from your AWS config;
• Log in — an SSO session expired; one click renews it (see AWS SSO and EKS).

Once a minute KubeDrive sends an anonymous request (/version) to each API server and looks at the AWS profiles and SSO sessions your kubeconfig uses. It does not log in or run credential plugins until you open a cluster.

Switch between Cards, List and Compact with the buttons at the top right.

Open a cluster

Click a cluster. KubeDrive connects with the context’s credentials and opens its pods — in all namespaces, unless the context names a namespace or your role may only list some of them. The card at the top of the sidebar opens the cluster’s Overview: nodes, pods and everything that needs attention.

Production clusters

Clusters whose name looks like production (prod, prd, production, live) are marked protected: the window gets a red frame and changes ask you to type the cluster’s name. Change it per cluster with the pencil on its card.