GitHub deployment workflow

[tmkarlsen]

The deployment to prod environments must be set up to give us better control

[Kamelon7]

Currently, test and staging environments are automatically restarted to pull the latest Docker image each time a new one is built. This process isn't desirable for production environments.

Suggestion:
Create a manual process that tags a specific image with a tag (e.g. "stable" or "prod"). We can either add the server restart step here, or solve that part in AWS.