Trying to run Django app on Gunicorn locally using Foreman

Question

I'm following Getting Started with Django on Heroku tutorial
When I'm trying to run Django app on the Gunicorn using Foreman, i get next error traceback:

09:23:33 web.1  | started with pid 7012  
09:23:34 web.1  | 2013-05-06 09:23:34 [7012] [INFO] Starting gunicorn 0.17.2  
09:23:34 web.1  | 2013-05-06 09:23:34 [7012] [INFO] Listening at: `http://0.0.0.0:5000` (7012)  
09:23:34 web.1  | 2013-05-06 09:23:34 [7012] [INFO] Using worker: sync  
09:23:34 web.1  | 2013-05-06 09:23:34 [7015] [INFO] Booting worker with pid: 7015  
09:23:34 web.1  | 2013-05-06 09:23:34 [7015] [ERROR] Exception in worker process:  
09:23:34 web.1  | Traceback (most recent call last):  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 485, in spawn_worker  
09:23:34 web.1  |     worker.init_process()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/workers/base.py", line 100, in init_process  
09:23:34 web.1  |     self.wsgi = self.app.wsgi()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/app/base.py", line 103, in wsgi  
09:23:34 web.1  |     self.callable = self.load()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/app/wsgiapp.py", line 25, in load  
09:23:34 web.1  |     return util.import_app(self.app_uri)  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/util.py", line 372, in import_app  
09:23:34 web.1  |     __import__(module)  
09:23:34 web.1  |   File "/Users/cosmicMan66/DjangoDev/hepcat_server/hepcat_server/wsgi.py", line 27, in <module>  
09:23:34 web.1  |     from django.core.wsgi import get_wsgi_application  
09:23:34 web.1  | ImportError: No module named django.core.wsgi  
09:23:34 web.1  | Traceback (most recent call last):  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/arbiter.py", line 485, in spawn_worker  
09:23:34 web.1  |     worker.init_process()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/workers/base.py", line 100, in init_process  
09:23:34 web.1  |     self.wsgi = self.app.wsgi()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/app/base.py", line 103, in wsgi  
09:23:34 web.1  |     self.callable = self.load()  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/app/wsgiapp.py", line 25, in load  
09:23:34 web.1  |     return util.import_app(self.app_uri)  
09:23:34 web.1  |   File "/usr/local/lib/python2.7/site-packages/gunicorn/util.py", line 372, in import_app  
09:23:34 web.1  |     __import__(module)  
09:23:34 web.1  |   File "/Users/cosmicMan66/DjangoDev/hepcat_server/hepcat_server/wsgi.py", line 27, in <module>  
09:23:34 web.1  |     from django.core.wsgi import get_wsgi_application  
09:23:34 web.1  | ImportError: No module named django.core.wsgi  
09:23:34 web.1  | 2013-05-06 09:23:34 [7015] [INFO] Worker exiting (pid: 7015)  
09:23:35 web.1  | 2013-05-06 09:23:35 [7012] [INFO] Shutting down: Master  
09:23:35 web.1  | 2013-05-06 09:23:35 [7012] [INFO] Reason: Worker failed to boot.  
09:23:35 web.1  | exited with code 3  
09:23:35 system | sending SIGTERM to all processes  
SIGTERM received

Procfile is located in the root directory of a project and contains:

web: gunicorn hepcat_server.wsgi

settings.py is located in hepcat_server directory

when I use $ python manage.py run_gunicorn
gunicorn starts successfully and I see default Django page

Answer 1

I had the same issue while following the Heroku setup page you were using. I found a solution on this page. I've copied it here for posterity:

In the root directory of your django project, create the Procfile file. then write this on it web: sh -c "cd djangoproject && gunicorn djangoproject.wsgi". then create another file called .env in the same location, write DJANGO_SETTINGS_MODULE=djangoproject.settings into it, this will help you setting the required env variable that cannot be set in djangoproject.wsgi module. run foreman start in order to test. What this does should be self explanatory giving the issue below.

When you deploy, after creating your application in heroku, set the env variable with this command heroku config:set DJANGO_SETTINGS_MODULE=djangoproject.settings, this is because the deployment fails before the file djangoproject/djangoproject/wsgi.py is run.

remember to replace "djangoproject" with the name of your django project.