asyncio and asyncio.gather

Detailed answer

Logic: async def coroutines await non-blocking IO. asyncio.gather runs coroutines concurrently; return_exceptions=True prevents one failure cancelling others. Complexity notes included in code section when present. Pros: High concurrency for IO-bound workloads on single thread. Cons: CPU-bound code blocks loop — use ProcessPoolExecutor; debugging async stack traces harder. Core: async def coroutines await non-blocking IO. asyncio.gather runs coroutines concurrently; return_exceptions=True prevents one failure cancelling others. Real-time example: FastAPI endpoint gathers user DB query + Redis cache + external HTTP in parallel. Pros: High concurrency for IO-bound workloads on single thread. Cons: CPU-bound code blocks loop — use ProcessPoolExecutor; debugging async stack traces harder. Common mistakes: Calling blocking requests.get inside async def; forgetting await on gather result. Best practices: Use httpx.AsyncClient; set timeouts; limit concurrency with Semaphore. Audience level: Tech Lead.

Full explanation

async def coroutines await non-blocking IO. asyncio.gather runs coroutines concurrently; return_exceptions=True prevents one failure cancelling others.

Code example

import asyncio

async def fetch(n: int) -> int:
    await asyncio.sleep(0.1)
    return n * n

async def main():
    results = await asyncio.gather(fetch(2), fetch(3), fetch(4))
    print(results)  # [4, 9, 16]

asyncio.run(main())

Practice code · python (view only · no execution)

import asyncio

async def fetch(n: int) -> int:
    await asyncio.sleep(0.1)
    return n * n

async def main():
    results = await asyncio.gather(fetch(2), fetch(3), fetch(4))
    print(results)  # [4, 9, 16]

asyncio.run(main())

Follow-up questions

• How would you test asyncio and asyncio.gather?
• What metrics prove asyncio and asyncio.gather is healthy in prod?
• How does asyncio and asyncio.gather change at 10× traffic?