Hi, I’m running some flows on prefect cloud and tasks are running sequentially instead of in parallel. The reason I decided to use prefect is to easily parallelize task execution. I am using

task.map

. What am I doing wrong?

The default executor is a sequential executor that does not parallelize. You need to set the LocalDaskExecutor.

flow.executor = LocalDaskExecutor()

Does it work also with

DaskExecutor

?

Yes if you have a cluster. Yes you need LocalDask or Dask to parallelize

when I do that and re-register the flow to the prefect-cloud I get `Skipped (metadata unchanged)`and still seeing sequential execution in the prefect cloud ui.

If you register using the CLI, you can force the change with

prefect register -f ..

thanks.