Svarog Forges

i.e. plain-vanilla async service blueprint for compute intensive tasks

 

The project's purpose was to tailor minimalistic async Python service template for CPU absorbing tasks. Here distributed message passing pattern is used where long computations are baked behind the curtain. The following have been accomplished until now:

• main service impl in FastAPI frame with all batteries included i.a. Swagger & OAuth2;
• worker realisation based on Celery (+Flower) and Redis as quee message broker;
• some compute intensive tasks examples i.e. fractals img generation / neural nets training..;
• backend dbase solution (Redis again?) for storing serialized tasks results;
• trivial static tasks monitoring webapp tailored with pure JS, Jinja & Html\CSS.

#QUICKTOUR:

# service deploy & run with docker-compose:
sudo docker compose down && sudo docker compose up -V --remove-orphans