The `restart django` task fails on first run. about mapstory

[jwood106]

Saw Tyler created this issue. Pasting a change I made locally in attempt to address this issue. Not convinced it is the solution, and may spark better idea.

Synopsis: add a 'start django' task and run both the existing 'restart django' and the new 'start django' tasks, allowing the provision to continue (instead of the current halt) if it fails. If django is already started, 'restart-django' task should succeed, if django is not already started, 'start django' task should succeed.

Changes to two files:

scripts/provision/admin.yml
(add use of start django)

 - name: gunicorn-app
   copy: src=files/gunicorn-app.sh dest=/srv/scripts/gunicorn-app.sh mode=650 owner=mapstory group=www-data
-  notify: [restart django]
+  notify: [start django, restart django]
   tags: [config, files, gunicorn]

 - name: supervisor-gunicorn
   copy: src=files/supervisor-gunicorn-django.conf dest=/etc/supervisor/conf.d
-  notify: [reload supervisor, restart django]
+  notify: [reload supervisor, start django, restart django]
   tags: [config, gunicorn]

scripts/provisions/handlers.yml
(add definition of start django, add ignore_errors to existing restart django)

- name: restart django
  supervisorctl: name=gunicorn-django state=started
  sudo: yes
  sudo_user: root
  notify: check django running
+ # jcw - allow it to fail, might not be started.
+ ignore_errors: True

+  - name: start django
+  supervisorctl: name=gunicorn-django state=started
+  sudo: yes
+  sudo_user: root
+  notify: check django running
+ # jcw - allow it to fail, might already be started.
+  ignore_errors: True

[garnertb]

This is fixed.