AsyncIO & Concurrency

setup setup python asyncio concurrency parallelism 1 min read

Python async/await, threads, processes, event loops, and synchronization

Run CPU-bound operations while waiting for IO-bound responses.

Concurrency vs Parallelism

Concurrency: Two lines, one cashier (takes turns). Works well in Python.
Parallelism: Two lines, two cashiers. Faces GIL in Python (single-threaded even with multiple threads).

Key Concepts

async: Marks function to run concurrently
await: Controls execution order (e.g., waiting for DB response)
gather: Runs functions in parallel

import asyncio
res1, res2 = await asyncio.gather(func1(args1), func2(args2))

When to Use What

AsyncIO: Managing many waiting tasks
Threads: Parallel tasks sharing data with minimal CPU use (IO-bound)
Processes: Maximizing performance on CPU-intensive tasks (runs across multiple cores)

Core Components

Event Loop

The core that manages and distributes tasks.

Coroutines

async def main():
    print("Start of main coroutine")

asyncio.run(main())

Synchronization

Lock: Mutual exclusion
Semaphore: Limited concurrent access
Event: Setter/waiter pattern