Don't miss your chance to take the Fabric Data Engineer (DP-700) exam on us!
Learn moreWe've captured the moments from FabCon & SQLCon that everyone is talking about, and we are bringing them to the community, live and on-demand. Starts on April 14th. Register now
When publishing a report from Power BI Desktop to the Service, the dataset is automatically refreshed if a scheduled refresh is configured. I propose adding an option to disable the automatic refresh on publish, so that publishing only updates the metadata/report, not the dataset.
Currently, when a dataset has scheduled refresh enabled, publishing a new version of a report triggers an immediate on-demand refresh in the Service. This behavior is widely observed in the community and is considered standard behavior for datasets with scheduled refresh enabled.
However, this creates several problems:
Unnecessary refreshes
If the dataset is already refreshed (e.g., refreshed locally in Desktop before publishing), there is no need to refresh it again.
Long waiting times
For large datasets, refresh can take 30–60 minutes or more, delaying availability of small report changes (layout, measures, RLS, etc.).
Operational inefficiency
In environments with frequent refresh schedules (daily or multiple times per day), triggering an additional refresh on every publish wastes compute resources and capacity.
Blocked deployment scenarios
Users often need to make quick report or security changes (e.g., RLS roles), but cannot finalize them until the refresh completes.
Publishing a report should not automatically trigger a dataset refresh when:
A scheduled refresh is already configured, and
The dataset does not require immediate data changes.
Instead, users should have control over the behavior.
Add a setting such as:
“Refresh dataset after publish” (On/Off toggle)
This could be configured:
Per dataset in the Service, or
As a checkbox in the publish dialog in Power BI Desktop
Default behavior could remain unchanged for backward compatibility, but users should be able to disable it.
Faster deployment of report changes
Reduced load on capacity and gateways
Better control over refresh timing
Improved user experience for enterprise BI teams
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.