Power BI Fabric Capacity Monitoring: Linking CU Consumption to Specific Reports

[SCUTARI_Santiag]

Hello,

We're looking for some guidance on optimizing our Power BI Fabric capacity monitoring, and we're hoping someone here might have experience with a similar challenge.

Our Setup: We heavily utilize Power BI shared semantic models (mostly organized by specific business use cases). These models serve as the data source for numerous reports across our organization.

The Challenge: When we analyze our Power BI Fabric Capacity Consumption metrics (using the Fabric Capacity Metrics app), the consumption data for Capacity Units (CUs) is primarily reported at the semantic model level.

This presents a significant challenge: we are unable to accurately attribute or link the CU consumption back to the individual reports that are querying and driving the usage on a particular semantic model. While we know which semantic model is consuming resources, we can't pinpoint which report is the primary driver of that consumption within that model.

Our Goal: Our objective is to gain more granular visibility into capacity usage, specifically to identify:

• Which reports are consuming the most CUs.
• Whether specific reports are experiencing performance issues related to capacity.
• How we can optimize individual reports to reduce overall capacity consumption.

Is there a direct or recommended method within Power BI Fabric (or via external tools/scripts) to track CU consumption by individual report, rather than just by semantic model?

Any insights, tips, or experiences you can share would be incredibly valuable. 

Thank you in advance for your time and expertise!

[Poojara_D12]

Hi @SCUTARI_Santiag 

You're not alone in facing this challenge—many organizations using Power BI Fabric encounter the same limitation when it comes to granular capacity monitoring, especially in environments with heavy reliance on shared semantic models. The Fabric Capacity Metrics app provides insights primarily at the semantic model level, which is helpful for broad analysis, but insufficient for pinpointing which specific reports are driving high Capacity Unit (CU) consumption. This lack of granularity makes it difficult to diagnose performance bottlenecks or optimize individual report usage.

Unfortunately, Power BI currently does not offer a built-in or direct method to track CU consumption per report within a semantic model. However, there are some recommended practices and workarounds you can consider to bridge this gap:

1. Power BI Activity Log + Query Logs: By combining the Power BI Activity Log (which captures user-level actions like report views and dataset queries) with the XMLA query logs or Query Diagnostics from Premium workspaces, you can correlate the frequency and intensity of report usage. While this won't directly translate to CUs, it gives you a proxy to identify which reports are likely contributing most to semantic model load.

2. Log Analytics Integration: If you're using Microsoft Purview or enabling Azure Monitor / Log Analytics integration with your Fabric capacity, you can collect query-level telemetry and refresh performance data, allowing deeper exploration of which reports or users are generating heavy queries.

3. Custom Telemetry: Implement telemetry in your reports using Power BI usage metrics, custom audit logging (e.g., through Power Automate flows that track report opens), or usage tagging techniques (such as embedding report metadata or unique user/session info in query parameters). You can then analyze usage patterns alongside CU data to approximate report-level impact.

4. Segment Shared Models: For very high-traffic semantic models, consider creating segmented versions (by business unit, report group, or department) to isolate CU usage and monitor more cleanly at a smaller scale.

5. Feature Requests and Feedback Loop: Microsoft is aware of this limitation, and you can submit or vote for this feature through the Power BI Ideas portal to push for per-report CU monitoring in future updates.

In summary, while direct report-level CU tracking is not natively available today, combining Activity Logs, Log Analytics, and smart usage telemetry can give you a reasonable level of insight. It's an area where careful architecture (e.g., segmenting shared models) and monitoring strategy can significantly improve visibility and capacity planning.

[v-achippa]

Hi @SCUTARI_Santiag,

Thank you for reaching out to Microsoft Fabric Community.

Thank you @Deku and @mohitkumawat for the prompt response.

As we haven’t heard back from you, we wanted to kindly follow up to check if the solution provided by the user's resolved your issue? or let us know if you need any further assistance.
If any response resolved your issue, please mark it as "Accept as solution" and click "Yes" if you found it helpful.

Thanks and regards,

Anjan Kumar Chippa

[v-achippa]

Hi @SCUTARI_Santiag,

We wanted to kindly follow up to check if the solution provided by the user's resolved your issue? or let us know if you need any further assistance.
If any response resolved your issue, please mark it as "Accept as solution" and click "Yes" if you found it helpful.

Thanks and regards,

Anjan Kumar Chippa

[v-achippa]

Hi @SCUTARI_Santiag,

As we haven’t heard back from you, we wanted to kindly follow up to check if the solution provided by the user's resolved your issue?
If any response resolved your issue, please mark it as "Accept as solution" and click "Yes" if you found it helpful.

Thanks and regards,

Anjan Kumar Chippa

[Deku]

If you use workspace monitoring or log analytics you can get all the execution metrics logs. These will include all Dax queries and their CPU usage and duration, per report and visual. While this doesn't correspond direct to CU, you can get a idea of distribution between reports

[mohitkumawat]

Hi @SCUTARI_Santiag ,

As per Microsoft Fabric Capacity Metrics app: This is the official and most comprehensive tool for monitoring your Fabric capacity consumption. so basically  it Identify top CU consumers (items and operations) as shown in image. 

for Optimizing Reports there are lot of factor but i highlights some of like:-, 

• Streamline your data model by importing only necessary data and optimizing data types.
• Prefer DAX measures over calculated columns to reduce model size and improve refresh times.
• Optimize visuals and report layout by limiting visuals per page and choosing efficient visual types.
• Implement strategic data refresh practices like incremental refresh and smart scheduling.
• Utilize performance monitoring tools such as Power BI Performance Analyzer, DAX Studio, and Tabular Editor.