Deploying an Existing Django App to Heroku on a Mac

Because the tutorial almost, but didn’t quite, work.

The Django tutorial Heroku provides ALMOST worked. I ironed out a couple crinkles when I went through the process myself so I thought I would share it with the world to save people some heartache. You should read this in conjunction with that tutorial to get your Django app up and running on Heroku.

First Things First

First, if you do not have Postgresql, you should get it because you will need it for Heroku. I was originally using an SQLite backend—this does not work on Heroku and you will need to download Postgresql. If you have Homebrew,

brew install postgresql

If not, install Homebrew and then install Postgresql. Then, in your Django App,

pip install virtualenv

virtualenv venv

source venv/bin/activate

*export CFLAGS=-Qunused-arguments export CPPFLAGS=-Qunused-arguments*

pip install django-toolbelt

Note the extra line. I ran into a terrible error during “pip install django-toolbelt” (mno-fused-madd MADNESS related to XCode Command Line Tools). The export CFlags line will save you from this torture.

Create a Procfile. This should be in our project folder (outer folder) but not within your app folder. In your Procfile,

web: gunicorn INSERTNAMEOFDJANGOPROJECT.wsgi

You should now be able to, in terminal:

foreman start

My foreman start merely said “web: 1: started pid #” rather than the “Starting gunicorn” promised by the tutorial. The default port is 5000 with foreman so you should be able to navigate to localhost:5000 and see your site.

Back to Heroku Config Stuff

As the tutorial states, run:

pip freeze > requirements.txt

This will generate a requirements.txt with all of your dependencies.

Setting up Postgresql database on Heroku

Create a Postgresql database at https://postgres.heroku.com/databases. When I created a database, it automatically generated a Heroku app with a generic name to go along with it. I was able to change the name of the Heroku app later to a more appropriate URL by visiting my apps dashboard. Back on the databases dashboard, click on the database name to view Host, Database Name, Username, Password, etc.

In your settings.py, replace your existing DATABASES field with the fields specified for the Heroku database created. Please note that these are not your Heroku username and password but rather the username and password provided for that specific Heroku generated database:

DATABASES = {
 ‘default’: {
 ‘ENGINE’: ‘django.db.backends.postgresql_psycopg2',
 ‘NAME’: ‘DATABASE_FROM_HEROKU’, 
 ‘USER’: ‘USERNAME_FROM_HEROKU’,
 ‘PASSWORD’: ‘PASSWORD_FROM_HEROKU’,
 ‘HOST’: ‘HOST_FROM_HEROKU’
 }
}

This step is not included in the tutorial Heroku provided but needs to be done to use their Postgresql backend.

Next, follow tutorial instructions by adding the djdatabaseurl lines and static asset configuration lines to your settings.py file.

# Parse database configuration from $DATABASE_URL
import dj_database_url
DATABASES['default'] = dj_database_url.config()

# Honor the 'X-Forwarded-Proto' header for request.is_secure()
SECURE_PROXY_SSL_HEADER = ('HTTP_X_FORWARDED_PROTO', 'https')

# Allow all host headers
ALLOWED_HOSTS = ['*']

# Static asset configuration
import os
BASE_DIR = os.path.dirname(os.path.abspath(__file__))
STATIC_ROOT = 'staticfiles'
STATIC_URL = '/static/'

STATICFILES_DIRS = (
 os.path.join(BASE_DIR, 'static'),
)

Note on Static Files:

Essentially Django pulls the used static files in project(outer)/app(inner)/static/ and collects and copies them to the project(outer)/staticfiles/ directory.

If this is set up correctly, you should be able to run

python manage.py collectstatic —noinput

while in virtualenv and have it collect the static files to the project(outer)/staticfiles/ directory. When visiting the site locally (localhost:8000) with “python manage.py runserver”, you should be able to access your static files at “localhost:8000/static/…”. In your templates, you can access these static files as such

 {% load staticfiles %}

<link href=”{% static “bootstrap/css/bootstrap.min.css” %}” rel=”stylesheet” type=”text/css” />

Back to Heroku Stuff

DO NOT follow the tutorial in adding “Cling” to wsgi.py. This did not work for me and only caused gunicorn to break. Simple leave wsgi.py alone and ensure that “foreman start” still runs as expected.

I did not change my wsgi.py at all—it is as follows:

import os
os.environ.setdefault(“DJANGO_SETTINGS_MODULE”, “APP.settings”)

from django.core.wsgi import get_wsgi_application
application = get_wsgi_application()

Last Mile

Create a .gitignore if you do not already have one:

venv
*.pyc
staticfiles

Hopefully you already have a git. If not,

git init

git add .

git commit -m "message"

heroku git:remote -a NAMEOFHEROKUAPP

Lastly,

git push heroku master

and

heroku ps:scale web=1

heroku run python manage.py syncdb

Set up your database as you see fit and run:

heroku open

Hopefully your Heroku Django app is up and running. If not, feel free to comment or reach out.

Menu

Because the tutorial almost, but didn’t quite, work.