Flyte enables production-grade orchestration for machine learning workflows and data processing created to accelerate local workflows to production.

Flyte

sample_map_test.py

Hi, I’m trying to write a unit test for a workflow that contains a `map_task`. This map_task has the `min_success_ratio=0.5`. I’ve tested the workflow in our dev cluster and it seems to work as I would expect. If enough tasks pass the `map_task` succeeds and the failed tasks are returned as empty.
However in unit testing it appears to ignore the `min_success_ratio` and if any task fails it immediately exits with an exception.
Here’s a small test to illustrate the issue

would you mind filing an issue for this?  it’s not supported at the moment.  i don’t know if this is super high priority for us though - how badly do you need this?

i think the point of the failure ratio is more to account for failures on a live backend.  with the understanding that local execution should pretty much always work

Thanks, it’s not critical, but happy to file an issue
I was trying to write a unit test for the workflow to ensure we had setup the map task as desired

I didn’t realize map_task called a different execute function for local vs remote execution