Forum Discussion
How do you currently audit a messy Power BI semantic model?
I'm exploring a tool idea to automatically scan a Power BI/Fabric workspace for duplicate/conflicting measures, undocumented DAX, and unused reports — instead of doing it by hand. Before I build much more, I wanted to ask people who actually deal with this: what does your current cleanup/audit process look like? Is it mostly manual? Any tools you already use for this? Genuinely trying to learn here, not pitching anything yet.
Hi siddhi1803
Thanks for reaching out to Fabric Community.
Few options are Tabular Editor's Best Practice Analyzer, extracting measures through
DMVs or INFO.VIEW.MEASURES(), or using Fabric notebooks with semantic-link-labs.
Make sure that the XMLA endpoint these tools rely on is Premium/PPU/Fabric only, not Pro even though the messiest sprawl is often in Pro, where everything is still manual. That’s why most governed setups I see are on Premium/Fabric.
Reference docs: Semantic model connectivity and management with the XMLA endpoint in Power BI - Microsoft Fabric | Microsoft Learn
Hope it helps!
5 Replies
- tayloramy
Super User
Hi siddhi1803,
Scanning for duplicate measures needs to be somewhat manual, there's no API endpoint that will expose that. You can use tools like Tabular Editor to connect to the XMLA endpoint for a workspace and do it though.
For unused reports, use the GetActivityEvents endpoint Admin - Get Activity Events - REST API (Power BI Power BI REST APIs) | Microsoft Learn
This will let you know what users do in the system.
Only 28 days of data is stored in the API, so I recommend setting up a job that will copy it to a lakehouse or warehouse nightly so you can maintain history.Thanks so much, this is really helpful — especially the Activity Events tip, I hadn't accounted for the 28-day retention limit. Quick follow-up if you don't mind: in your experience, is duplicate/inconsistent measure detection something teams actually struggle with at scale, or is it more of a minor annoyance? And just curious — are the environments you work with typically on Premium/Fabric capacity, or mostly Pro
- v-csrikanth
Community Support
Hi siddhi1803
Thanks for reaching out to Fabric Community.
Few options are Tabular Editor's Best Practice Analyzer, extracting measures through
DMVs or INFO.VIEW.MEASURES(), or using Fabric notebooks with semantic-link-labs.
Make sure that the XMLA endpoint these tools rely on is Premium/PPU/Fabric only, not Pro even though the messiest sprawl is often in Pro, where everything is still manual. That’s why most governed setups I see are on Premium/Fabric.
Reference docs: Semantic model connectivity and management with the XMLA endpoint in Power BI - Microsoft Fabric | Microsoft Learn
Hope it helps! - v-csrikanth
Hi siddhi1803
We would like to inquire whether have you got the chance to check the solutions provided above in commiunity to resolve the issue. We hope the information provided helps to clear the query. Should you have any further queries, kindly feel free to contact the Microsoft Fabric community.
Thank you. - v-csrikanth
Hi siddhi1803
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. And, if you have any further query do let us know.
Thank you.
