Pipeline Parameters Reverting to Dataflow Defaults – How to Confirm Actual Applied Values?

[mh20221111]

Hi everyone,

I’m seeing an unexpected behavior with parameters in my data pipeline.

I set parameters in the data flow, and then from the pipeline activity I pass values for those parameters and run the pipeline. However, when I open the pipeline this morning, the parameters configured in the activity have been overwritten—now they show the *initial/default values* from the data flow instead of the values I had set in the activity.

If I run the pipeline *without saving* it, it still uses the previously configured parameter values (so it works), but then I can’t easily confirm which parameter values are currently applied to the pipeline, since the activity screen doesn’t reflect them.

What’s the best way to check which parameter values are currently set/applied to the pipeline (even if the UI has reverted to the data flow defaults)?

Any advice would be appreciated.

[mh20221111]

@v-achippa @RossEdwards @blopez11 

Thank you for your message. I will keep an eye on Issue ID 1856 going forward.

[v-achippa]

Hi @mh20221111,

Thank you for reaching out to Microsoft Fabric Community.

Thank you @blopez11 and @RossEdwards for the prompt response.

As we haven’t heard back from you, we wanted to kindly follow up to check if the solution provided by the user's for the issue worked? or let us know if you need any further assistance.

Thanks and regards,

Anjan Kumar Chippa

[blopez11 · Accepted solution]

We ran into this same issue and reported to Microsoft back in early June.  We were informed yesterday that a fix may be rolling out on 6/24.  Please see https://support.fabric.microsoft.com/known-issues/?product=Data%2520Factory&active=true&fixed=false&... which I believe describes the issue.

[RossEdwards]

Normally with a Semantic Model you could set a rule for each of the steps in your deployment pipeline to be assigned a different parameter value when the model is deployed through that pipeline.  My understanding is that functionality is not available to dataflows and the parameters set in the environment is carried across as part of the  deploy action.

[mh20221111]

I added a screenshot of the problematic part to the question. From the Data Pipelines screen, it looks as if the parameters for Gen2 Dataflow are being set to the same values directly. However, originally the pipeline was configured with different values, and when this pipeline is actually executed, it runs using the values that were configured in the pipeline.

Thanks.

[RossEdwards]

The problem is that dataflows have never really supported parameters properly.  You can add a paramater to your dataflow but unlike a semantic model, there is no method to update those parameters in the settings menu.  You have "Edit" the dataflow to make those changes.

For this reason, i believe the pipelines cannot update the parameters.