Deploy a Python (Django) web app container to Azure App Service (in progress)
This Python web app is a restaurant review application using the Django framework. The app is intended to be used as a container running in Azure App Service with a connection to MongoDB such as Azure Cosmos DB API for MongoDB.
Following the tutorial Containerized Python web app on Azure, you can deploy this web app to Azure, hosted in a fully managed Azure App Service. Azure managed identity enables the App Service to pull container images from an Azure Container Registry. MongoDB connection info is passed to the code through environment variables.
If you need an Azure account, you can create on for free.
A Flask sample application with similar functionality is at Flask Container Web App.
Run and deploy options
Here are some ways you can run the sample web app in this repository.
| Scenario | As-is repo code | Containerized app |
|---|---|---|
| Local environment | Run repo code in virtual environment with requirements.txt. Set environment variables in shell before running. For an example, see Configure a Python virtual environment. | Build image from repo and run locally in Docker container. Pass environment variables in Docker CLI command or with VS Code task definition 1. This scenario is covered in Containerized Python web app on Azure. |
| Azure App Service | Deploy repo code to App service. Set environment variables as App Service configuration settings. See note about subpath. 2 | Deploy container to Web App for Containers. Build image locally or in Azure and push to container registry like Azure Container Registry. Configure App Service to pull from registry. Set environment variables as App Service configuration settings. This scenario is covered in Containerized Python web app on Azure. |
| Azure Container Apps | n/a | Build image locally or in Azure and push to container registry like Azure Container Registry. Create a Container App with deployment from the registry. Configure environment variables for the container. |
(1) The .vscode directory settings.json and tasks.json are configured either for a MongoDB local connection or an Azure Cosmos DB connection. The tasks and templates in .vscode are only used when using Visual Studio Code locally.
(2) This app was designed to be containerized and run on App Service. If you want to deploy to App Service without containerizing it, then be sure to set the subpath to the azureporject folder, which contains the manage.py file. Do that by:
- In VS Code: F1 or CTRL + SHIFT + P to open the command palette.
- Type "Preferences: Open Workspace Settings" and select to open.
- In the Settings search field, enter "@id:appService.deploySubpath appser".
- Set the subpath as "azureproject".
Environment variables
The sample code requires the following environment variables passed in as described in the scenario table above.
CONNECTION_STRING=<connection-string>
DB_NAME=restaurants_reviews
COLLECTION_NAME=restaurants_reviews
For a local MongoDB instance, the connection string is of the form mongodb://127.0.0.1:27017. An Azure Cosmos DB API for MongoDB connections string is of the form mongodb://<server-name>:<password>@<server-name>.mongo.cosmos.azure.com:10255/?ssl=true&<other-parameters>.
Requirements
The requirements.txt has the following packages:
| Package | Description |
|---|---|
| Django | Web application framework. |
| gunicorn | Gunicorn ‘Green Unicorn’ is a Python WSGI HTTP Server for UNIX. |
| pymongo | The PyMongo distribution contains tools for interacting with MongoDB database from Python. |
| requests | An HTTP library |
| whitenoise | Static file serving for WSGI applications, used in the deployed app. This package is used in the azureproject/settings.py file. |