I built a pipeline that runs a foreach loop that triggers a notebook to refresh Power BI semantic models. It's pretty simple - I pass parameters into the notebook in the first cell and then use the parameters to refresh like this:
import sempy.fabric as fabric
fabric.refresh_dataset(
workspace = WorkspaceName,
dataset = SemanticModel
)
print(WorkspaceName, SemanticModel)
Note that I'm experiencing a service bug where the Semantic Model Refresh activity isn't available (I submitted a ticket).
This refresh appears to be working - the last refresh date is updated and the refresh is logged. However, the reports themselves aren't updated. If I refresh manually, they work. The difference between the two that I see is this.
Scheduled refresh shows:
Via Enhanced Api
# Type Start End Duration Status Error
| 1 | Data | 5/8/2024, 12:42:02 PM | 5/8/2024, 12:42:03 PM | 00:00:00.875 | Completed |
Manual refresh shows:
# Type Start End Duration Status Error
| 1 | Data | 5/8/2024, 3:17:39 PM | 5/8/2024, 3:17:49 PM | 00:00:09.517 | Completed |
| 1 | Query Cache | 5/8/2024, 3:17:49 PM | 5/8/2024, 3:17:49 PM | 00:00:00.203 | Completed |
It appears that the query cache step is necessary to refresh the report visuals, but how do I do this with sempy.fabric?
