I am trying the newly released run_if option for tasks. I have created a workflow with one single leaf task that depends on multiple parent tasks and its run_if is set to AT_LEAST_ONE_SUCCESS.
The job definition is deployed with dbx and I would expect it to run without errors.
Current Behavior
The job fails when started with "The job run failed because task dependency types are temporarily disabled". However, the following procedure can fix it:
manually change the cluster definition from the UI (e.g., change the runtime) and save
manually revert the change
now the job works
Steps to Reproduce (for bugs)
Just deploy with dbx a workflow which uses the run_if option for one of its tasks and try to run it.
Expected Behavior
I am trying the newly released
run_if
option for tasks. I have created a workflow with one single leaf task that depends on multiple parent tasks and itsrun_if
is set toAT_LEAST_ONE_SUCCESS
.The job definition is deployed with
dbx
and I would expect it to run without errors.Current Behavior
The job fails when started with "The job run failed because task dependency types are temporarily disabled". However, the following procedure can fix it:
Steps to Reproduce (for bugs)
Just deploy with
dbx
a workflow which uses therun_if
option for one of its tasks and try to run it.Context
Related forum thread: https://community.databricks.com/t5/get-started-discussions/the-job-run-failed-because-task-dependency-types-are-temporarily/m-p/39166#M672
I'm not sure if this is a
dbx
bug or it's a problem with REST APIs.Your Environment