Cluster
Start a local k3d cluster or import an existing kubeconfig to manage any Kubernetes environment.
Start a local k3d cluster or import an existing kubeconfig, then work from the Kubernetes section.
Start a local cluster with k3d
Use k3d when you want a disposable local cluster for development and testing.
Open the Kubernetes section
If no cluster is configured, Dockerman shows a Start a Cluster button.
Click Start a Cluster
Dockerman downloads the k3d binary on first use.
Wait for the cluster to come up
The local cluster usually starts in about 10 seconds on a modern machine.
Start using Kubernetes
The namespace switcher populates and workload lists become available.
Dockerman downloads the k3d binary on first use. Allow outbound HTTPS or install the binary yourself before you start.
Import an existing cluster
Import a kubeconfig when you already have a cluster and want Dockerman to manage it directly.
Click Import Cluster
Open the import flow from the Cluster page.
Select a kubeconfig file
Dockerman parses the file and lists its contexts. Pick the context you want to import; the current flow does not promise separate pre-filled server URL and user-binding fields.
Verify the connection
Dockerman probes the API server before it saves the cluster.
Switch clusters
Use the cluster switcher at the top of the Kubernetes section to move between imported clusters and any local k3d cluster.
- Cluster selection is independent of the active Docker host. Switching to a different K8s context does not touch your Docker connection, and vice versa.
- Your current Kubernetes page is preserved when switching clusters where the same resource exists; the namespace selector syncs with the new cluster.
- After an app restart Dockerman reconnects to the last cluster and restores deep pages such as the YAML editor or a CRD instance you were viewing.
Delete a cluster
Use the action that matches the cluster type.
- Delete from Dockerman to forget an imported kubeconfig while leaving the remote cluster running.
- Tear down the local k3d cluster to remove the cluster and its data.
Tearing down a k3d cluster is irreversible. You lose every workload and persistent volume created inside it.
Cluster version and status
The Cluster page shows the API server version, connection state, and node information. Kubernetes Node objects expose kubelet and container-runtime versions, not a per-node kube-proxy version.
Nodes, PersistentVolumes, and Namespaces
Each of these gets a dedicated in-app page under the Kubernetes section. Reach them from the K8s sidebar:
- Nodes — node list with conditions, capacity, allocatable resources, labels, kubelet version, container runtime, and scheduled Pods.
- PersistentVolumes — cluster-scoped PV list with phase, claim binding, reclaim policy, and capacity. See Config & Storage → PersistentVolumes for the full reference.
- Namespaces — create, delete, and inspect namespaces directly. The sidebar namespace selector syncs with this page.
Cluster events
Open the dedicated Events browser to filter by namespace, type (Normal/Warning), reason, and involved object before you narrow to a single resource.