Asyncio.create_task() 和 asyncio.gather() 依次运行

Question

I have an API, which takes a long time to do. So I want to split it into many smaller jobs, then run them in parallel and wait for the result before sending the response.

My snip code:

@app.post("/data")
async def get_tsp_events:
    query = [foo, bar, foo, bar]
    tasks = [asyncio.create_task(do_work(param1, param2)) for query in queries]
    events = await asyncio.gather(*tasks)
    return events

async def do_something(arg1, arg2):
    log("Start time")
    # This take a lot of times
    events = [event for event in range(10000000000)]
    log("End time")
    return events

As I see, all task is run in sequence just like normal code (without using asyncio.create_task() and asyncio.gather ) I'm new in Python and my question is:

• I'm wrong or not ? Where ?
• There are any solution can help me ?

Thank you all

Answer 1

1. In fact, async and await didn't introduce real parallel, they works when you do some underlying "asynchronous" operations like aiohttp , aiofile . Nevertheless, they only bring concurrency for IO-bound tasks.

2. If you want real parallel for CPU-bound tasks, use multiprocessing .