Start existing cluster during up when stopped

README.md

@@ -237,6 +237,11 @@ running the full pipeline:
 ./jeannie fix-debian-docker-root
 ```
 
+When `./jeannie up` sees an existing tracked kubeadm control plane but
+the Kubernetes API is down, it starts the saved runtime before applying cluster
+changes. Use `./jeannie rebuild-cluster` only when you intentionally
+want to destroy and recreate the cluster.
+
 ## Validation
 
 Useful checks after a rebuild:

README.md.tmpl

@@ -237,6 +237,11 @@ running the full pipeline:
 ./{{ main_script }} fix-debian-docker-root
 ```
 
+When `./{{ main_script }} up` sees an existing tracked kubeadm control plane but
+the Kubernetes API is down, it starts the saved runtime before applying cluster
+changes. Use `./{{ main_script }} rebuild-cluster` only when you intentionally
+want to destroy and recreate the cluster.
+
 ## Validation
 
 Useful checks after a rebuild:

jeannie

@@ -3789,6 +3789,7 @@ up() {
     if [[ -z "${LAB_CLUSTER_VAR_FILE:-}" ]]; then
         prepare_cluster_worker_var_file true
     fi
+    ensure_existing_cluster_started_for_up
     run_tofu_stack "bootstrap/cluster"
     doctor_versions_report
     run_tofu_stack "bootstrap/platform"
@@ -4006,6 +4007,45 @@ start_cluster() {
     echo "Kubernetes runtime start requested. Use 'kubectl get nodes -o wide' to watch readiness."
 }
 
+kubernetes_api_reachable() {
+    kubectl --kubeconfig "${KUBECONFIG_PATH}" get --raw=/readyz >/dev/null 2>&1
+}
+
+cluster_control_plane_tracked() {
+    tofu_state_has_resource "bootstrap/cluster" "null_resource.kubeadm_control_plane"
+}
+
+wait_for_kubernetes_api() {
+    local timeout_seconds="${1:-180}"
+    local elapsed=0
+
+    until kubernetes_api_reachable; do
+        if ((elapsed >= timeout_seconds)); then
+            echo "Kubernetes API did not become reachable after ${timeout_seconds}s." >&2
+            return 1
+        fi
+        sleep 5
+        elapsed=$((elapsed + 5))
+    done
+}
+
+ensure_existing_cluster_started_for_up() {
+    if ! cluster_control_plane_tracked; then
+        echo "No tracked kubeadm control plane found; bootstrap/cluster will create one."
+        return 0
+    fi
+
+    if kubernetes_api_reachable; then
+        echo "Existing Kubernetes control plane is reachable."
+        return 0
+    fi
+
+    echo "Existing Kubernetes control plane is tracked but API is down; starting cluster runtime..."
+    start_cluster
+    wait_for_kubernetes_api "${LAB_CLUSTER_START_WAIT_SECONDS:-180}"
+    echo "Existing Kubernetes control plane is reachable after start."
+}
+
 status_section() {
     printf '\n== %s ==\n' "$1"
 }