Get started

Set up a local instance for development


CoReport is a project in development. The project and this documentation are updated regularly, and these notes may have changed since the last time you read them.

Clone the codebase:

git clone

The project requires that you have Docker installed; you’ll need:

cd into the covid19-report directory and build the application:

docker-compose build

This will create images based on the services described in the docker-compose.yml file, including the web service, The Celery services, the Postgres database and any others.

Migrate the database:

docker-compose run web python migrate

This runs the Django migrate inside a web container.

Launch the application:

docker-compose up

It will take a few moments to start up all the containers, and then you will be able to reach the application at http://localhost:8000 (this is configured in docker-compose.yml). You should see an unstyled login page.

In order to generate the styling for the site, along with all its other frontend components, they will need to be built manually. This requires Yarn, a Node package, to be installed. (It is beyond the scope of this documentation to describe that process.)

In a new terminal session in the project directory, run:

yarn install --pure-lockfile

This installs the components that the project’s frontend requires. Then:

yarn start

This launches the frontend; the frontend assets will be compiled and the served on port 8090. Yarn watches the frontend source and will re-compile the assets on any change.

When you refresh the login page, the styling should now appear.

Prepare the project for cloud deployment

The CoReport project is ready for deployment on the Divio cloud management platform, and the project already contains the components required for this. All your repository requires in order to test in a cloud deployment is to associate it with a Divio cloud project.

Using the Divio Control Panel (you will need an account, it’s free), create a new project.

Select the defaults:

  • Platform: Python

  • Type: Django

You can use Divio’s Git server (also the default).

Hit Skip in the Subscription view.

After a few moments, you’ll be presented with its dashboard, which will look something like this:

'New project in the Dashboard'

Install the Divio CLI: pip install divio-cli.


divio login

This will fetch a token from

Upload your public key to

Get your project’s slug from the Control Panel:

'Project slug'

Or you can see it by listing your projects:

divio project list

Set up the project:

divio project setup <project slug>

Once the divio project setup command has finished pulling down the repository, you will find a file in the newly-created directory:: .aldryn. The file contains the slug and id of the cloud project, and is what the Divio CLI uses to associate the local project with its counterpart on the cloud.

Move this file into the CoReport covid19-report project.

We need to do a similar thing with Git. In the local Divio project, find its Git URL:

git remote show origin

In the covid19-report project, add a new remote called divio, using this URL, for example:

git remote add divio

Your covid19-report project is now associated with the Divio cloud project you just created. You can dispose of the local project that was created by the divio project setup command; it’s no longer needed.

You can now interact with both the origin remote (for example, to pull new updates) and push your changes to the divio remote (to test it on the cloud).

Working with the project

Basic workflow:

git add <files>  # to stage changes for commit
git commit  # to commit changes
git push divio  # push changes to your cloud project
divio project deploy  # deploy the cloud Test server

Also useful:

divio project pull [or push] db  # copy the database to/from the cloud project

You can access the Test and Live sites of your cloud project via the Control Panel. See the Divio developer handbook for a quick guide to using the Divio CLI.

Whenever you pull or make changes that will require rebuilding the local project (for example, changing re-run:

docker-compose build