SecFit

SecFit (Secure Fitness) is a hybrid mobile application for fitness logging.

Port mapping shenanigans

The port mapping in this application is a bit strange to enable hosting multiple stages (development/staging/production) with multiple containers for each, without needing to handle multiple DNSs.

We use static variables combine these to create unique ports for each container:

PREFIX is always defined as 2 and is found in the .env file as PORT_PREFIX=2. This is needed in order for the full port to be on the form 2XXXX, to avoid crashing with established ports.

GROUP_ID is defined by your group number on BlackBoard, found in the .env file as GROUP_ID=XX. This is always 2 numbers, so groups 1-9 have leading zeros, i.e., group 3 has GROUP_ID=03.

STAGE is defined as:

development: 1
staging: 2
production: 3

The related variables can be found in the .env file as DEV_POSTFIX=1, STAG_POSTFIX=2, and PROD_POSTFIX=3.

CONTAINER_ID is defined as:

backend: 4
frontend: 5
gateway: 6

The related variables can be found in the .env file as BACKEND_POSTFIX=4, FRONTEND_POSTFIX=5, and GATEWAY_POSTFIX=6.

The ports section of each compose file (docker-compose.dev.yml, docker-compose.stag.yml, and docker-compose.yml) define which images are accessible from outside the isolated Docker network. Currently, all images are available in development and staging, while only the gateway is available in production.

To make ports available from outside the server, you'll need to either modify firewall/routing on the server or (as we have elected to do for this project) use a backward proxy. This is defined through the ./nginx_template.txt file, which, when used with the $ envsubst < nginx_template.txt command, replaces the variables with the environment variables set by the .env file (how to do this properly is left as part of task 1 in exercise 2).

The default setup to make the application available from the outside is to listen on ports PREFIX + GROUP_ID + STAGE + 7 (and port 80 for production), mapping to their related gateways PREFIX + GROUP_ID + STAGE + 6.

One note on this is to make sure the proxy_set_header Host $http_host; is set with $http_host and not $host (as it normally is) in order to forward the port being used (otherwise, it assumes port 80 based on the default HTTP protocol).

Port examples

Group Number	Stage	Image	Port
04	development	backend	20414
21	staging	gateway	22126
00	production	frontend	NA (not available)

Deploy with Docker

Prerequisites:

Docker

Git

Windows hosts must use Education or more advanced versions to run Docker
Download: https://innsida.ntnu.no/wiki/-/wiki/English/Microsoft+Windows+10

Install:

In your github, fork this repository (https://git.ntnu.no/mathialm/secfit)

$ git clone https://git.ntnu.no/<username>/secfit
$ cd secfit

Run:

In this project, we have three docker-compose files you can use to set up the containers.

docker-compose.dev.yml is used to make the backend, frontend, and gateway containers with all ports open. These are intended to be used as development containers (i.e., enabling CI/CD and updating as code is pushed).

docker-compose.stag.yml is used to make the backend with a different configuration (in ./backend/tests/Dockerfile.test). This is intended to be used for testing your code and should only run when tests are run. You can expand this to also set up the frontend and gateway if implemented tests require it.

docker-compose.yml is used to make the backend, frontend, and gateway containers, with only the ports on the gateway open. These are intended to be used as production containers and should, as such, ensure persistent volumes and proper migration where needed.

Verify compose configuration file:

$ docker compose -f <compose_file> --verbose config

Build and run containers defined in configuration file

$ docker-compose -f <compose_file> up --build -d

Using GitHub actions

Managing GitHub actions is done from your repository's Actions tab.

From here, you can add, modify, and read the logs from all workflows.

The workflow files are stored in the .github/workflows/ folder.

Documentation and tutorials for GitHub Actions can be found at https://docs.github.com/en/actions

Technology

deployment - Docker/Docker Compose/GitHub actions
web - Nginx
database - SQLite 3
backend - Django 4 with Django REST framework
frontend - HTML5/CSS/JS, React 4
authentication - JWT

Backend code and structure

secfit/ django project folder containing the project modules
- <application_name>/ - generic structure of a Django application
  - migrations/ - directory containing django migrations enabling mostly automatic modifications to your data models, see https://docs.djangoproject.com/en/5.1/topics/migrations/
  - admin.py - file containing definitions to connect models to the Django admin panel
  - apps.py - defined the application for this directory
  - forms.py - definitions of forms. Used to render html forms and verify user input
  - models.py - contains data models
  - parsers.py - contains custom parsers for parsing the body of HTTP requests
  - permissions.py - contains custom permissions that govern access
  - settings.py - Contains important settings at the application and/or project level
  - serializers.py - contains serializer definitions for sending data between backend and frontend
  - tests/ - contains tests for the module. See https://docs.djangoproject.com/en/2.1/topics/testing/ for more
  - urls.py - contains the mapping between URLs and views
  - views.py - Controller in MVC. Methods for rendering and accepting user data
- media/ - directory for file uploads
- comments/ - application handling user comments and reactions
- secfit/ - The projects main module containing project-level settings
- tests/ - Container Docker config for test and test_template
- users/ - application handling users and requests
- workouts/ - application handling exercises and workouts
- db.sqlite3 - SQLite database file
- manage.py - entry point for running the project.
- requirements.txt - Python requirements

Local setup

It's recommended to have a look at: https://www.djangoproject.com/start/ Just as important is the Django REST guide: https://www.django-rest-framework.org/

Create a virtualenv https://docs.python-guide.org/dev/virtualenvs/

Django

Installation with examples for Ubuntu. Windows and OSX is mostly the same

Fork the project and clone it to your machine.

Setup and activation of virtualenv (env that prevents python packages from being installed globally on the machine)

Navigate into the project folder, and create your own virtual environment

Install python requirements

$ pip install -r requirements.txt

Migrate database

$ python manage.py migrate

Create superuser

Create a local admin user by entering the following command:

$ python manage.py createsuperuser

Only username and password is required

Start the app

$ python manage.py runserver

Add initial data

You can add initial data by going to the URL the app runs on locally and adding /admin to the URL.

Add some categories and you should be all set.

Assumptions

The file .env would normally not be uploaded together with the code since that is bad practice allowing for sensitive data exposure. The file is however essential to run the application, and needs to stay where it is. If any other secret values needs to be stored 'outside' the code repository, this is where it should be put. If you use a different file, it should still be added to the repository, but you should state it here as an assumption.

julielsu/secfit

About

Resources

Stars

Watchers

Forks

Releases

Packages 0

Languages

Packages