Memory Refresh Failure in Cloud

[shuhn1229]

Hi all,

I have a large report that refreshes just fine in desktop but I am getting the following error when trying to refresh in serivce: 

[ValidateMarkupTags][ccon]DataSource.Error: AnalysisServices: [ccon]You have reached the maximum allowable memory allocation for your tier. Consider upgrading to a tier with more available memory. Technical Details: RootActivityId: 005e3b50-c6f5-4b6d-989e-7317cc7cde67 Date (UTC): 4/5/2024 1:21:12 PM[/ccon]. DataSourceKind = [ccon]AnalysisServices[/ccon].

Is this an error coming from Anaalysis Service or is this due to my workspace not having enough memory? I am in premium. Thanks.

[Anonymous · Accepted solution]

Hi @shuhn1229 

@amitchandak  Thank you for your prompt reply!

I would like to add the following points:

The error is indeed related to the Analysis Services, which Power BI uses under the hood for data modeling and refresh operations. When you're working in Power BI Desktop, the memory constraints are based on your local machine's resources, which is why you might not encounter this issue locally. However, when you move to refresh in the Power BI service, your operations are subject to the memory limits of the tier you're subscribed to.

Before considering an upgrade, try to optimize your report and the underlying data model. This includes removing unnecessary columns, aggregating data at a higher level, and using more efficient DAX expressions. For more information on optimizing your Power BI reports, please refer to the:Burstable capacity - Microsoft Fabric | Microsoft Learn

 If optimization doesn't alleviate the issue, you may need to upgrade to a higher tier with more memory. To do this, sign into the Home - Microsoft Azure, select the Microsoft Fabric service to see your capacities, and follow the steps to scale your capacity. Detailed instructions can be found Scale your Fabric capacity - Microsoft Fabric | Microsoft Learn.

Best Regards,

Jayleny

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[shuhn1229]

I resolved this by reducing the size of a few queries

[amitchandak]

@shuhn1229 , you need to review, what data and transformations you are using. If you are premium capacity, it should able to load unless it is huge data, with no incremental setting or some complex power query transformation