They will only actually return failure if they fail. This with the rate limit may work nicely for API-based tasks. Though it would be dangerous to assume a rejected task meant that the rate limit was hit. So retries would be safer.
It's pretty primitive but these are just meant to act as soft rate limiters or to help improve reliability in systems where there may be outages or other reasons a function may not pass initially but might pass on a second run.
They will only actually return failure if they fail. This with the rate limit may work nicely for API-based tasks. Though it would be dangerous to assume a rejected task meant that the rate limit was hit. So retries would be safer.
It's pretty primitive but these are just meant to act as soft rate limiters or to help improve reliability in systems where there may be outages or other reasons a function may not pass initially but might pass on a second run.