Dataflow loads in preview, but not in actual dataset?

MB_Dem · 2023-02-27T14:41:41+00:00

There was an error rendering this rich post.

MB_Dem

Any troubleshooting tips for us? Our data flow's outputs:

Work when we run a preview

When we run it, the datasets give us an "Unable to load this DataSet" error. (The second dataset is my "test" output where I kept moving it around the flow, trying to identify if a specific part of the dataflow was breaking the final output.)

Find more posts tagged with

Accepted answers

MB_Dem

FYI Turns out it was a glitch in the matrix. After the person recreated the dataflow, using the EXACT SAME datasets/tiles/joins/etc, worked. ¯\_(ツ)_/¯

All comments

GrantSmith

does the user you’re logged in as have permissions to those output datasets?

MB_Dem

Yes. Myself and the person who created them both have Author roles, and full permissions to all the datasets referenced.

MichelleH

@MB_Dem By default any dataflow output datasets are owned by the user who owns the dataflow, even if both users have access to all the input datasets. The owner of the dataflow will need to share both output datasets with you.

MB_Dem

To clarify, the dataflow was created by the same person who also gets the "Unable to load dataset error."

I've directed them to create a brand new dataflow, recreate the work inside the original, and see if that resolves the problem. (I'm pretty sure we experienced this glitch before, and mysteriously a new dataflow worked out the kinks. ¯\_(ツ)_/¯ )

I'll provide an update if the new dataflow solves the problem.

MB_Dem

FYI Turns out it was a glitch in the matrix. After the person recreated the dataflow, using the EXACT SAME datasets/tiles/joins/etc, worked. ¯\_(ツ)_/¯

McSQL

@MB_Dem I'm glad you have a way to resolve the issue. if the original dataflow is still in existence, I would suggest sending to the support team as it may help fix the bug. Would be nice to have the root cause fixed so that you and others don't run into this in the future.