fix: bound CSI rollout wait time and add failure diagnostics

ansible/roles/csi/tasks/main.yml

@@ -14,15 +14,50 @@
   changed_when: true
 
 - name: Wait for CSI controller rollout
-  command: kubectl rollout status deployment/hcloud-csi-controller -n kube-system
+  command: kubectl rollout status deployment/hcloud-csi-controller -n kube-system --timeout=30s
   register: csi_controller_rollout
   until: csi_controller_rollout.rc == 0
   retries: 30
   delay: 10
   changed_when: false
 
+- name: Show CSI controller status on failure
+  command: kubectl -n kube-system get deployment hcloud-csi-controller -o wide
+  register: csi_controller_deploy_status
+  changed_when: false
+  failed_when: false
+  when: csi_controller_rollout is failed
+
+- name: Show CSI controller pods on failure
+  command: kubectl -n kube-system get pods -l app=hcloud-csi-controller -o wide
+  register: csi_controller_pods_status
+  changed_when: false
+  failed_when: false
+  when: csi_controller_rollout is failed
+
+- name: Show CSI driver logs on failure
+  command: kubectl -n kube-system logs deployment/hcloud-csi-controller -c hcloud-csi-driver --tail=120
+  register: csi_driver_logs
+  changed_when: false
+  failed_when: false
+  when: csi_controller_rollout is failed
+
+- name: Fail with CSI controller diagnostics
+  fail:
+    msg: |
+      CSI controller rollout failed.
+      Deployment status:
+      {{ csi_controller_deploy_status.stdout | default('n/a') }}
+
+      Pods status:
+      {{ csi_controller_pods_status.stdout | default('n/a') }}
+
+      hcloud-csi-driver logs:
+      {{ csi_driver_logs.stdout | default('n/a') }}
+  when: csi_controller_rollout is failed
+
 - name: Wait for CSI node daemonset rollout
-  command: kubectl rollout status daemonset/hcloud-csi-node -n kube-system
+  command: kubectl rollout status daemonset/hcloud-csi-node -n kube-system --timeout=30s
   register: csi_node_rollout
   until: csi_node_rollout.rc == 0
   retries: 30