SQL Database Connector, PowerBI Desktop and Throttling

[bhalicki]

Hi all,

Wondering if anyone can tell me if the SQL Database connector in PowerBi Desktop is subject to API throttling limits?  I have a rather large model I am trying to refresh and no matter what I do (filtering rows first, only importing 5 rows, buffering, increasing CPU/Memory options, 10G backbone), the model refresh gets stuck 'Evaluating'.  Problems seem to have started around after I updated in April.  The Dataverse connector was pathetic, so I switched to the SQL connector which worked for a while, but has now become unresponsive as well.

I wish there was some way of skipping the model evaluation stage, so I can just save, publish to the service and apply the changes there.

I am connecting to the Dataverse.

Kind regards,

Ben.

[v-nmadadi-msft]

Hi @bhalicki 

As we haven’t heard back from you, we wanted to kindly follow up to check if the suggestions  provided by the community members for the issue worked. Please feel free to contact us if you have any further questions.

Thanks and regards

[v-nmadadi-msft]

Hi @bhalicki 

May I check if this issue has been resolved? If not, Please feel free to contact us if you have any further questions.

Thank you

[v-nmadadi-msft]

Hi @bhalicki 

I wanted to check if you had the opportunity to review the information provided. Please feel free to contact us if you have any further questions.

Thank you.

[v-nmadadi-msft · Accepted solution]

Hi @bhalicki 
Thanks for reaching out to the Microsoft fabric community forum.

The maximum size for semantic models imported into the Power BI service is 1 GB. These semantic models are heavily compressed to ensure high performance. In addition, in shared capacity, the service places a limit of 10 GB on the amount of uncompressed data that is processed during refresh. This limit accounts for the compression, and therefore is higher than the 1-GB maximum semantic model size. Semantic models in Power BI Premium aren't subject to these limits. If refresh in the Power BI service fails for this reason, reduce the amount of data being imported to Power BI and try again.
Addittionally Scheduled refreshes for imported semantic models time out after two hours. This time-out is increased to five hours for semantic models in Premium workspaces. If you encounter this limit, consider reducing the size or complexity of your semantic model, or consider refactoring the large semantic model into multiple smaller semantic models.

I hope this information helps. Please do let us know if you have any further queries.
Thank you

[Shahid12523]

1. Disable Auto Data Preview
- Go to Options → Query Editor → Uncheck “Allow data preview to download in the background”
2. Force Query Folding
- Use native SQL or minimal transformations to ensure folding happens at source.
- Avoid Table.Buffer, Table.Combine, or custom columns early in the query.
3. Use DirectQuery Temporarily
- Switch to DirectQuery mode for testing—this skips model evaluation and may help isolate the issue.
4. Publish First, Then Apply Changes
- Create a minimal model, publish to Power BI Service, then apply transformations via Dataflows or Service-side transformations