Appreciate any HELP! Consider the following code snippet:

import aiohttp
import asyncio
import time

async def fetch(session, url):
    async with session.get(url) as response:
        return await response.text()

async def main():
    url = "<http://127.0.0.1:5000/simulate>"
    tasks = []
    async with aiohttp.ClientSession() as session:
        for _ in range(1000):  # Adjust number of requests
            tasks.append(fetch(session, url))
        await asyncio.gather(*tasks)

if __name__ == "__main__":
    asyncio.run(main())

• This code will run

fetch

concurrently 1000 times, all in a single thread, asynchronously. • After I explored Prefect

.submit() /.map()

, it became obvious that Prefect concurrency model is multithreading, which means adding a huge overhead launching and switching between threads in I/O-bound use cases like the one above. My question: • How can I convert

fetch

to a

@task

and

main

to

@flow

(to benefit from retrying and tracking infrastructure) and at the same time executing the task async in a single thread? • I understand it is possible to use Dask or Ray as taks executers, and it seems these libraries independently support async execution in a a single thread (much like what you expect in vanilla python), so how can I set them up with Prefect to work like that?

hi Mo - tagging me is not the best way to get help, i dont have enough time for everyone to tag me. if you want async python, then you would write async tasks and flows that use asyncio normally if there's some problem or discussion you want to have, please open a discussion: https://github.com/PrefectHQ/prefect/discussions

OK thank you!