Idea:
When trying to implement dynamic notebook execution in Microsoft Fabric Pipelines, the current system has significant limitations:
In the Pipeline Notebook activity, we can only select notebooks from a fixed dropdown list.
To make notebook execution dynamic (e.g., based on parameters or metadata), we must pass the Notebook Artifact ID — not the Notebook Name.
If we try to use a dynamic expression or pass the notebook name instead of the Artifact ID, the pipeline throws an error.
This makes it difficult to build reusable and dynamic notebook mappings, such as those based on configurations, metadata tables, or conditional logic.
Proposed Solution:
Allow dynamic reference to notebooks by name in the Notebook activity within pipelines — similar to how other systems (like Azure Data Factory) support dynamic activity references.
Benefits:
Enables metadata-driven and modular pipeline design
Reduces tight coupling between pipeline definitions and specific notebook IDs
Improves maintainability, especially in enterprise scenarios
Makes pipelines easier to reuse and scale across workspaces and teams
This enhancement would greatly improve flexibility and support dynamic orchestration patterns in Fabric Pipelines.
Thanks for considering this!
on:
Close, Apply, and Refresh All
nschachter
on:
Provide information on Gen2 dataflows in listing o...
miguel
on:
Allow user to modify applied steps in a Dataflow w...