Asyncio vs Threads: Avoiding I/O Blocking in Modern Python

前言 / Introduction

在 Python 專案的開發過程，或多或少會碰到需要平行運行的時候，而大多的資料只說使用 asyncio 就可以加速，但實際上執行時，卻還是一樣的緩慢甚至更糟，那或許你也踩到了錯誤使用線程的坑。

在這篇文章，將討論下面幾點:

• Why I/O-bound and CPU-bound tasks behave differently
• The right way to use asyncio
• Why requests.get() blocks async functions-

This sets the foundation for understanding Python 3.13 and 3.14 improvements later in the series.

I/O-Bound vs CPU-Bound: The Real Difference

在討論 asyncio or threads 之前，我們需要先搞清楚甚麼樣的任務類型是你面對的

• If your app is fetching APIs, reading files, or handling sockets → it’s I/O-bound.
• If it’s crunching data or doing math → it’s CPU-bound.

本篇文章為了更清楚描述，接下來會專注在 I/O-bound 的討論。

Why Asyncio Exists (and Why Threads Are Not Enough)

雖然 Python 的 Thread 已經能夠很好的執行 I/O 任務，Interpreter 會在等待 I/O 的時候去釋放 GIL。但 asyncio 還可以更有效率的使用 single-threaded event loop，優化 Task 的執行。

For Example where asyncio shines:

• HTTP calls
• File operations
• WebSockets & streaming
• Database queries
• Message queues

But asyncio only works if your code is non-blocking.

❌ The Classic Mistake: Using requests.get() in async def

有效率的使用，是在沒有踩到 asyncio 的坑，比較常見的錯誤案例，就是在 async def 中混用會造成 I/O blocking 的函式。

import asyncio
import requests

async def bad_fetch(url):
    resp = requests.get(url)  # ❌ This blocks the entire event loop
    return resp.text

async def main():
    await asyncio.gather(
        bad_fetch("https://httpbin.org/delay/3"),
        bad_fetch("https://httpbin.org/delay/3")
    )

asyncio.run(main())

即使已經使用 asyncio.gather()，但結果還是 sequential execution。 requests.get() 會導致進程被 blocking， 所以 event loop 將會被鎖住，直到 request 完成。

✅ Correct Approach: Use an Async HTTP Client (aiohttp)

那避免踩坑的方式，就是正確使用相關的第三方 Library 。 以上面的例子，使用 aiohttp才是正確的方法。

import asyncio
import aiohttp

async def good_fetch(url):
    async with aiohttp.ClientSession() as session:
        async with session.get(url) as resp:
            return await resp.text()

async def main():
    await asyncio.gather(
        good_fetch("https://httpbin.org/delay/3"),
        good_fetch("https://httpbin.org/delay/3")
    )

asyncio.run(main())

• ✅ Now both requests happen concurrently
• ✅ Event loop stays responsive
• ✅ No threads needed

When to Choose Asyncio vs Threads for I/O

Key Takeaways

• Asyncio only works if your I/O calls are non-blocking
• requests.get() blocks the event loop → use aiohttp or executor workaround
• Standard library can do async sockets, but it’s low-level
• Threads still work for I/O, but async scales better
• This sets the stage for why free-threading and new interpreters matter later