Deploying Python apps

Python is a programming language that lets you work more quickly and integrate your systems more effectively. You can learn how to use Python and see almost immediate gains in productivity and lower maintenance costs.


Python 2.7, 3.6 and 3.7 are available on our platform. You can use Git to deploy your application.

Create an application

Refer to the page Deploy an application on Clever Cloud.


You are granted to install external libs. As you can do on your workstation you can easily use pip and requirements.txt.

For example to install Flask and various libs you have to create a file /requirements.txt :

psycopg2>=2.7 --no-binary psycopg2

We recommend using psycopg2>=2.7 --no-binary psycopg2 to avoid wsgi issues.

Cache dependencies

You can cache dependencies to avoid the installation at each deployment. Define the CC_CACHE_DEPENDENCIES variable to true to activate it. If dependencies have changed between deployments, the cache will be updated. Remove the environment variable or set it to false to disable this feature.

If you have any question, feel free to contact our support.

Configuring the application

Python apps can be launched in a variety of ways. You can specify how to start your application (for instance which module to run) by setting environment variables (in the "Environment variables" section of the console).

Select your module

To select which module you want to start, use the CC_PYTHON_MODULE variable.


The module (without .py) must be importable, i.e. be in PYTHONPATH. For example with Flask, it's gonna be the name of your main server file followed by your Flask object: server:app for example if you have a file at the root of your project with a Flask app object inside.

Basically, you should just point to a WSGI capable object.

Celery apps

Note: Please note that Celery support is not available yet for gunicorn.

We also support celery apps out of the box. To deploy a celery app, use the CC_PYTHON_CELERY_MODULE variable:


You can also activate beat with CC_PYTHON_CELERY_USE_BEAT=true and provide a given log dir for celery with CC_PYTHON_CELERY_LOGFILE="/path/to/logdir".

The CC_PYTHON_CELERY_LOGFILE path is relative to the application's path.

Beware of timezones with Celery Beat!

There is a bug in recent versions of Celery. You need to add the CELERY_TIMEZONE = 'UTC' environment variable. The bug is documented here:

Select the python backend

Currently, we support uwsgi and gunicorn for python backends. To select one, set the CC_PYTHON_BACKEND environment variable with either uwsgi or gunicorn.

If not specified, the default backend is uwsgi.

Using the Gevent loop engine

Whether you use uwsgi or gunicorn, you can enable the Gevent loop engine.

To do so, add the CC_PYTHON_USE_GEVENT environment variable to your application, with the true value.

Manage your static files

To enable Nginx to serve your static resources, you have to set two environment variables.

STATIC_FILES_PATH: should point to a directory where your static files are stored.

STATIC_URL_PREFIX: the URL path under which you want to serve static files (e.g. /public)

Also, you are able to use a Filesystem Bucket to store your static files. Please refer to the File System Buckets section.

Note: the path of your folder must be absolute regarding the root of your application.

Note: setting the STATIC_URL_PREFIX to / will make the deployment to fail.


Here is how to serve the static files, the test.png being the static file:

├── <app_root>
│   ├──
│   ├── static
│   │   └── test.png
│   └── requirements.txt

Using the environment variables:


The test.png file will be accessed under: https://<domain.tld>/public/test.png

Choose Python version

The default version of python on Clever Cloud is 2.7. If you want to use python 3.x instead, create an environment variable CC_PYTHON_VERSION equal to either 3 (which will default to the most up-to-date version), 3.6 or 3.7.

Note: the version is a number, do not use quotes. values allowed are 2, 2.7, 3, 3.6, 3.7.

Choose Pip version

The default version of pip on Clever Cloud is 19. If you want to use pip 9 instead, create an environment variable like this:


uWSGI, Gunicorn and Nginx configuration

uWSGI, gunicorn and nginx settings can be configured by setting environment variables:


  • HARAKIRI: timeout (in seconds) after which an unresponding process is killed. (Default: 180)
  • WSGI_BUFFER_SIZE: maximal size (in bytes) for the headers of a request. (Defaut: 4096)
  • WSGI_POST_BUFFERING: buffer size (in bytes) for uploads. (Defaut: 4096)
  • WSGI_WORKERS: number of workers. (Defaut: depends on the scaler)
  • WSGI_THREADS: number of threads per worker. (Defaut: depends on the scaler)


  • GUNICORN_WORKER_CLASS: type of worker to use. Default to sync. Available workers
  • CC_GUNICORN_TIMEOUT: gunicorn timeout. Defaults to 30


  • NGINX_READ_TIMEOUT: a bit like HARAKIRI, the response timeout in seconds. (Defaut: 300)
  • ENABLE_GZIP_COMPRESSION: "on|yes|true" gzip-compress the output of uwsgi.
  • GZIP_TYPES: the mime types to gzip. Defaults to text/* application/json application/xml application/javascript image/svg+xml.

Nginx configuration

Nginx settings can be configured further in clevercloud/http.json. All its fields are optional.

  • languages: configure a default language and redirections
  • error_pages: configure custom files for error pages
  • force_https: automatically redirect HTTP traffic to HTTPS
  • aliases: set up redirections
  • charset: force a specific charset
    "languages": {
        "default": {"rewrite": "en"},
        "fr": {"rewrite": "en"}
    "error_pages": {
        "404": "path/to/page"
    "force_https": true,
    "aliases": {
        "/path": "redirection"
    "charset": "latin-1"

Environment injection

Clever Cloud can inject environment variables that are defined in the dashboard and by add-ons linked to your application.

The access to these variables is simple: just get them as you would with any environment variable:

import os

uWSGI asynchronous/non-blocking modes

To enable uWSGI asynchronous mode, you can use these two environment variables:


We support execution of a single goal. Usually, this would be to execute custom tasks after the installation of dependencies. To execute a goal, you can define the environment variable PYTHON_SETUP_PY_GOAL="build" (or any other goal).

The goal will be launched after the dependencies from requirements.txt have been installed. tasks

We support execution of multiple tasks.

You can declare the tasks in clevercloud/python.json:

{ "deploy": { "managetasks": [ "migrate" ]}}

The tasks are launched after the dependencies from requirements.txt have been installed.

Git Deployment

You will need git on your computer to deploy via this tool. Here is the official website of Git to get more information:

Setting up your remotes

  1. The "Information" page of your app gives you your git deployment URL. It looks like this: git+ssh://<your_app_id>.git. Copy it in your clipboard.

  2. On your computer, go into your application repository. If you didn't already track your app with git, start by typing:

     $ git init
  3. Then, use the git remote command to add the deploy URL:

     $ git remote add <name> <your-git-deployment-url>
  4. The last step is to push your application:

     $ git push <name> master


The remote branch on Clever Cloud is ALWAYS master. If your local branch is not "master", use this syntax:

git push < name > yourbranch:master
Edit me on GitHub