Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

<@ULVA73B9P> which is best practice?


- Fail-fast by completion time (earliest failure raises immediately)

from prefect.futures import as_completed

futs = [task.submit(i) for i in range(10)]
for f in as_completed(futs):
    f.result()  # raises as soon as a failing future completes


- Resolve a collection (may block in traversal order, not completion order)
from prefect.futures import resolve_futures_to_results

futs = [task.submit(i) for i in range(10)]
results = resolve_futures_to_results(futs)  # raises when it reaches a failed future


- Wait for all, then raise if any failed (no fail-fast)

from prefect import futures

futs = [task.submit(i) for i in range(10)]
done, _ = futures.wait(futs)
for f in done:
    f.result()  # raises on the first failed future you iterate over

you shouldn't ever need to use `resolve_futures_to_results`, that's an internal util

this
```futs = [task.submit(i) for i in range(10)]
results = resolve_futures_to_results(futs)  # raises when it reaches a failed future```
should be written like this
```results = my_task.map(range(10)).result()```