Business framing: Async programming lets services handle many IO-bound tasks (requests, DB calls, websockets) efficiently. This notebook introduces asyncio fundamentals, a Pyodide-safe event-loop demo, when to choose async vs threads, and safe patterns for idempotency and cancellation.

Learning objectives¶

• Understand event loop, coroutine, task, and await.

• Compare async (single-threaded concurrency) vs thread pools for IO-bound work.

• Write simple asyncio code that is deterministic and safe to run in-browser.

Pyodide-safe asyncio demo (deterministic)¶

Visual intuition: event loop vs thread pool¶

Caption: Async uses a single event loop to multiplex coroutines; threads use OS threads for parallel blocking IO.

MCQ¶

• Q: For a high-concurrency web client that performs many short HTTP requests, which model usually gives the best throughput with lowest overhead?

  • A) Multiprocessing

  • B) Thread pool

  • C) Asyncio

  • (Answer: C)

Exercises¶

1. Modify simulated_io to sometimes raise a recoverable exception; add a retry wrapper using asyncio.sleep between retries.

2. Implement a simple cancellation test: start a long-running coroutine and cancel it after a short timeout, observing cleanup behavior.

3. (Stretch) Compare asyncio vs ThreadPoolExecutor on a simulated blocking function by measuring elapsed time for N tasks.