Forum Discussion
Power BI DataModel refresh getting failed after 24-25GB Size with Power BI Premium Per User Capacity
Issue:
The Power BI refresh start giving error "Data source error: The operation was throttled by Power BI because of insufficient memory. Please try again later." A common observation is when the Dataset size reach to 20GB it will start giving this error and once it grows to 24GB it will no longer get refreshed. Unable understand either the issue is from source DB or in Power BI capacity.
Details for Background Understanding:
We have a Data Model in which 3 tables are set for incremental refresh, rest are the full load dimension tables with only 100 record each. Incremental Tables growth rate is high. We have subscribed for Power BI Premium Per User License. The source system is MSSQL Database.
Our refresh frequency is after every hour. Incremental refresh is set as 3 months archive and incremental review window is 1 day.
Following settings are also enabled.
1. Large semantic Model storage format.
2. Query Scale-out.
Issue is after couple of days when Data Model start growing, we start facing the following error in our scheduled and on demand refreshes.
Error: Data source error: The operation was throttled by Power BI because of insufficient memory. Please try again later.
Sometimes it get refreshed after getting 1 failure sometimes till fails for multiple time and we have to manually cancel the refresh and re-initiate the refresh.
Note: The user is dedicated for that perticular PBI report. Only this report is refreshing. We can't say resources are being ultilized by any other report.
According to the Power BI Pricing the size of the dashboard and its capacity it can grow upto 100 GB.
Following is the snapshot of dataset.
Hi bilalrana
No, Power BI does not provide any interface to see live progress of an ongoing dataset refresh. In both Power BI Premium Per User (PPU) and Power BI Service generally, you can only see:
- That a refresh is in progress (rotating icon on the dataset),
- And the status (success or failure) once it finishes.
Unfortunately, there is no way to see real-time refresh progress, percentage complete, or which step is currently running. What you can do practically is, monitor the dataset’s Refresh History page to see when it started and when it finishes. Set up refresh failure alerts in Power BI Service, so you know immediately if something goes wrong (but again, no live progress). And final note is that Power BI simply doesn’t offer detailed refresh progress tracking only start/end status.
Please go through the Microsoft Official document for your reference.
Monitor Power BI data refresh - official docs
Data refresh in Power BI - Power BI | Microsoft Learn
If this response helps, consider marking it as “Accept as solution” and giving a “kudos” to assist other community members.
Thank's,
Akhil.
6 Replies
Hi bilalrana
No, Power BI does not provide any interface to see live progress of an ongoing dataset refresh. In both Power BI Premium Per User (PPU) and Power BI Service generally, you can only see:
- That a refresh is in progress (rotating icon on the dataset),
- And the status (success or failure) once it finishes.
Unfortunately, there is no way to see real-time refresh progress, percentage complete, or which step is currently running. What you can do practically is, monitor the dataset’s Refresh History page to see when it started and when it finishes. Set up refresh failure alerts in Power BI Service, so you know immediately if something goes wrong (but again, no live progress). And final note is that Power BI simply doesn’t offer detailed refresh progress tracking only start/end status.
Please go through the Microsoft Official document for your reference.
Monitor Power BI data refresh - official docs
Data refresh in Power BI - Power BI | Microsoft Learn
If this response helps, consider marking it as “Accept as solution” and giving a “kudos” to assist other community members.
Thank's,
Akhil.Hi bilalrana
You're using Power BI Premium Per User, which allows up to 100 GB model size, but during refresh, it only gets about 25 GB of memory.
Your dataset grows to ~24 GB, and refresh fails because it's crossing that memory limit during processing, not after.Move to Power BI Premium Capacity (P1 or higher)
- It gives more memory (up to 100 GB+ during refresh).
- Solves the throttling issue permanently.
If this report is business-critical and growing, Premium Capacity is the right long-term fix.
If you must stay on PPU for now, do these 3 things.
- Reduce Archive Period, change from 3 months to 1 or 2 months in incremental refresh, keeps the dataset size lighter.
- Remove Unused Columns, especially long text like ErrorMessage, every column adds memory load during refresh.
- Split Heavy Tables, create separate tables for historical and current data, only apply incremental refresh to current data.
________________________________________________________________________________________________________________________
If this response helps, consider marking it as “Accept as solution” and giving a “kudos” to assist other community members.Regards,
Akhil.Hi Anonymous
Thanks for the reply, I am currently using Power BI Premium Per User - Reserved which is SKU: PP3. Do we have any interface or anything which can tell us about the current progress of the refresh, as we get the status once the refresh is completed. It waste the time.
Hi bilalrana ,
I’m following up to confirm that all items discussed have been addressed. If there’s anything else you need, I’d be happy to assist.
Regards,
Akhil.Hi bilalrana ,
Just checking in once more were you able to review the details shared regarding Power BI dataset refresh monitoring? If you still have questions or need further clarification, feel free to share. Happy to assist further.
Thanks,
Akhil.Hi bilalrana ,
It’s been a little while, so I just wanted to check in one last time. Were you able to review the details shared regarding Power BI dataset refresh monitoring?
Since we haven’t heard back, and following the Microsoft Fabric Community Forum guidelines, we’ll go ahead and close this thread for now.
If you're still exploring solutions or need help with anything else in the future, please feel free to create a new post we’re always here to help.
Regards,
Akhil
