Properly recreate systemd unit on container update

roles/changedetection/vars/main/defaults.yml

@@ -1,3 +1,3 @@
 ---
 changedetection_username: changedetection
-changedetection_version: 0.45.21
+changedetection_version: 0.45.23

roles/podman-container/handlers/main.yml

@@ -1,7 +1,7 @@
 ---
 - name: Reload systemd (daemon-reload)
   become: true
-  become_user: "{{ actual_username }}"
+  become_user: "{{ container_user }}"
   ansible.builtin.systemd_service:
     daemon_reload: true
     scope: user

roles/podman-container/tasks/main.yml

@@ -37,6 +37,13 @@
       containers.podman.podman_container:
         name: "{{ podman_container_name }}"
         state: absent
+    - name: Make sure the systemd unit file isn't present
+      ansible.builtin.file:
+        path: "/home/{{ container_user }}/.config/systemd/user/container-{{ podman_container_name }}.service"
+        state: absent
+      notify: Reload systemd (daemon-reload)
+    - name: Flush handlers
+      ansible.builtin.meta: flush_handlers
     - name: Map volumes to Podman accepted list
       ansible.builtin.set_fact:
         volumes: "{{ volumes + ['/home/' + container_user + '/' + item.name + ':' + item.mnt] }}"
@@ -60,16 +67,9 @@
       notify: Reload systemd (daemon-reload)
     - name: Flush handlers
       ansible.builtin.meta: flush_handlers
-    # On restarts this seems to be a bit buggy. Some manual waiting might be necessary and running:
-    # systemctl --user daemon-reload && systemctl --user start {{ service_name }}
-    # The retry also doesn't work as of yet. Seems that doing a daemon-reload and then refreshing the task seems to work
     - name: Enable the newly created systemd service for user
       ansible.builtin.systemd:
         name: "container-{{ podman_container_name }}.service"
         state: started
         enabled: true
         scope: user
-      retries: 3
-      delay: 3
-      register: result
-      until: "'Error.EBUSY' not in result.msg"