Infrastructure includes following:
- MongoDB;
- mongo-express — web-client for MongoDB (available in /mongo-express/ or on the port 8081 if local);
- Portainer — web-client for Docker (available in /portainer/ or on the port 9000 if local).
Service API follows JSON:API specification. Read more here.
Requests are serving by nginx and proxying to gunicorn's workers. Acceptable content type — application/vnd.api+json.
Endpoints:
- /api/songs/?page[cursor]=&page[size]=&sort= — List of songs with cursor-based pagination;
- /api/songs/avg_difficulty/?level= — Average difficulty, level is optional;
- /api/search/?message= — Case-insensitive search by title and artist;
- /songs/:id/rate/ — Rate song by ID;
- /songs/:id/rating/ — Ratings (highest, lowest, average) of song by ID.
Create *.env files for every group from .env.d/.env.example, fill the values.
To keep data from MongoDB and Portainer create volumes via
docker volume create music_mongo and docker volume create music_portainer.
The only command to launch the prepared infrastructure is docker-compose up -d.
To launch the infrastructure locally execute docker-compose -f docker-compose.dev.yaml up -d.
Local launching doesn't use nginx, so you can start the service by executing python main.py.
Don't forget to change DB host from
mongotolocalhostin app.env
Execute ./init.sh to initialize database with test data and to create required indexes.
To run tests execute pytest inside root folder.
To measure coverage execute coverage run -m pytest and then coverage report.
Every new blueprint that is going to connect to MongoDB MUST inherit
MongoResourceBlueprintto processObjectIdcorrectly
DO NOT commit *.env files!
After any *.env update, all touched containers must be recreated via docker-compose up -d to apply new config.
Search every error in containers logs by command docker-compose logs and service_name at the end.
Also, don't forget to monitor containers' health by Portainer.