This recipe assumes that you have a Pyramid application setup using a Paste
INI file, inside a package called
myapp. This type of structure is found in
pyramid_starter scaffold, and other Paste scaffolds (previously called
project templates). It can be easily modified to work with other Python web
applications as well by changing the command to run the application as
Step 0: Install Heroku¶
Install the heroku gem per their instructions.
Step 1: Add files needed for Heroku¶
You will need to add the following files with the contents as shown to the
root of your project directory (the directory containing the
You can autogenerate this file with the following command.
$ pip freeze > requirements.txt
requirements.txt file, you will probably have a line with your
project's name in it. It might look like either of the following two lines
depending on how you setup your project. If either of these lines exist,
1project-name=0.0 2 3# or 4 5-e git+git@xxxx:<git username>/xxxxx.git....#egg=project-name
You can only use packages that can be installed with pip (e.g., those on
PyPI, those in a git repo, using a git+git:// url, etc.). If you have any
that you need to install in some special way, you will have to do that in
run file (see below). Also note that this will be done for every
instance startup, so it needs to complete quickly to avoid being killed by
Heroku (there's a 60-second instance startup timeout). Never include
editable references when deploying to Heroku.
Procfile with the following command.
$ echo "web: ./run" > Procfile
run with the following command.
#!/bin/bash set -e python setup.py develop python runapp.py
Make sure to
chmod +x run before continuing. The
develop step is
necessary because the current package must be installed before Paste can
load it from the INI file.
If using a version greater than or equal to 1.3 (e.g. >= 1.3), use the
1import os 2 3from paste.deploy import loadapp 4from waitress import serve 5 6if __name__ == "__main__": 7 port = int(os.environ.get("PORT", 5000)) 8 app = loadapp('config:production.ini', relative_to='.') 9 10 serve(app, host='0.0.0.0', port=port)
For versions of Pyramid prior to 1.3 (e.g. < 1.3), use the following for
1import os 2 3from paste.deploy import loadapp 4from paste import httpserver 5 6if __name__ == "__main__": 7 port = int(os.environ.get("PORT", 5000)) 8 app = loadapp('config:production.ini', relative_to='.') 9 10 httpserver.serve(app, host='0.0.0.0', port=port)
We assume the INI file to use is named
production.ini, so change the
runapp.py as necessary. The server section of the INI will
be ignored as the server needs to listen on the port supplied in the OS
Step 2: Setup git repo and Heroku app¶
Navigate to your project directory (directory with
setup.py) if not
already there. If your project is already under git version control, skip to
the "Initialize the Heroku stack" section.
Inside your project's directory, if this project is not tracked under git, it
is recommended yet optional to create a good
.gitignore file. You can get
the recommended python one by running the following command.
$ wget -O .gitignore https://raw.github.com/github/gitignore/master/Python.gitignore
Once that is done, run the following command.
$ git init $ git add . $ git commit -m "initial commit"
Step 3: Initialize the Heroku stack¶
$ heroku create --stack cedar
Step 4: Deploy¶
To deploy a new version, push it to Heroku.
$ git push heroku master
Make sure to start one worker.
$ heroku scale web=1
Check to see if your app is running.
$ heroku ps
Take a look at the logs to debug any errors if necessary.
$ heroku logs -t
Tips and Tricks¶
The CherryPy WSGI server is fast, efficient, and multi-threaded to easily
handle many requests at once. If you want to use it you can add
pastescript to your
setup.py:requires section (be sure to re-run
pip freeze to update the requirements.txt file as explained above) and
runapp.py to look like the following.
1import os 2 3from paste.deploy import loadapp 4from paste.script.cherrypy_server import cpwsgi_server 5 6if __name__ == "__main__": 7 port = int(os.environ.get("PORT", 5000)) 8 wsgi_app = loadapp('config:production.ini', relative_to='.') 9 cpwsgi_server(wsgi_app, host='0.0.0.0', port=port, 10 numthreads=10, request_queue_size=200)
Heroku add-ons generally communicate their settings via OS environment variables. These can be easily incorporated into your applications settings as show in the following example.
1# In your pyramid apps main init 2import os 3 4from pyramid.config import Configurator 5from myproject.resources import Root 6 7def main(global_config, **settings): 8 """ This function returns a Pyramid WSGI application. 9 """ 10 11 # Look at the environment to get the memcache server settings 12 memcache_server = os.environ.get('MEMCACHE_SERVERS') 13 14 settings['beaker.cache.url'] = memcache_server 15 config = Configurator(root_factory=Root, settings=settings) 16 config.add_view('myproject.views.my_view', 17 context='myproject.resources.Root', 18 renderer='myproject:templates/mytemplate.pt') 19 config.add_static_view('static', 'myproject:static') 20 return config.make_wsgi_app()