Task: Sharing Python Objects Across Multiple FastAPI Workers
Task: Sharing Python Objects Across Multiple FastAPI Workers: a task in Terminal-Lego-15k (Harbor dataset). When running FastAPI in production with multiple workers (using uvicorn or gunicorn), each worker process creates its own instance of Python objects. This means that state stored in memory…
The task
When running FastAPI in production with multiple workers (using uvicorn or gunicorn), each worker process creates its own instance of Python objects. This means that state stored in memory is not shared between workers, leading to inconsistent behavior.
Part of PrimeIntellect/Terminal-Lego-15k.