fix: increase k3s readiness timeout and emit diagnostics on failure

ansible/roles/k3s-server/tasks/main.yml

@@ -38,10 +38,35 @@
   command: "{{ (k3s_primary | default(false)) | ternary('kubectl get nodes', 'systemctl is-active k3s') }}"
   register: k3s_ready
   until: k3s_ready.rc == 0
-  retries: 30
+  retries: 120
   delay: 10
   changed_when: false
 
+- name: Show k3s service status on failure
+  command: systemctl status k3s --no-pager
+  register: k3s_status
+  changed_when: false
+  failed_when: false
+  when: k3s_ready is failed
+
+- name: Show recent k3s logs on failure
+  command: journalctl -u k3s -n 80 --no-pager
+  register: k3s_journal
+  changed_when: false
+  failed_when: false
+  when: k3s_ready is failed
+
+- name: Fail with k3s diagnostics
+  fail:
+    msg: |
+      k3s failed to become ready on {{ inventory_hostname }}.
+      Service status:
+      {{ k3s_status.stdout | default('n/a') }}
+
+      Recent logs:
+      {{ k3s_journal.stdout | default('n/a') }}
+  when: k3s_ready is failed
+
 - name: Copy kubeconfig to default location for root
   file:
     src: /etc/rancher/k3s/k3s.yaml