async-concurrency.md

Python by Example: Async Concurrency

Run multiple coroutines at the same time with asyncio.gather(). It starts all of them, waits for each to finish, and returns their results in order. Use this for parallel I/O—e.g., fetching several URLs. Concurrency here is cooperative: only one coroutine runs at a time, but they overlap while waiting.

What you'll learn:

• asyncio.gather() for concurrent execution
• Collecting results from multiple coroutines

import asyncio


async def fetch(id, delay):
    print(f"Task {id} starting")
    await asyncio.sleep(delay)
    print(f"Task {id} done")
    return id


async def main():
    # Run concurrently and collect results
    results = await asyncio.gather(
        fetch(1, 1),
        fetch(2, 0.5),
        fetch(3, 2),
    )
    print("Results:", results)


asyncio.run(main())

All three tasks start; task 2 finishes first (0.5s), then task 1 (1s), then task 3 (2s). Results are returned in the order you passed the coroutines.

To run this program:

$ python source/async-concurrency.py
Task 1 starting
Task 2 starting
Task 3 starting
Task 2 done
Task 1 done
Task 3 done
Results: [1, 2, 3]

Tip: Use asyncio.gather() when you have independent I/O tasks. For more control, use asyncio.create_task() and manage tasks yourself.

Try it: Add a fourth task with a different delay and observe the order of completion.

Source: async-concurrency.py

Next: Async Queues