top of page

Semantic model storage modes across an entire tenant

Autorenbild: Brunner_BIBrunner_BI

Measure Killer provides an interesting piece of information that is difficult to obtain through other means: tenant level storage modes of all semantic models in your tenant. This means we we will look at all models in Premium/Fabric and Pro workspaces. We can also include models in Personal workspaces (only possible for tenant admins).


Finding Composite- and Direct Query models helps us understand how Power BI is being used across the organization.


To get this information, start by selecting one of the two Tenant Analysis modes in Measure Killer, depending on whether you are a Fabric Admin or a Power Platform Admin in your organization or not:


Limited Tenant Analysis and Tenant Analysis modes in Measure Killer's starting window.
Limited Tenant Analysis and Tenant Analysis modes in Measure Killer's starting window.

This guide focuses on the Tenant Analysis mode, but the step-by-step is very similar in the Limited Tenant Analysis mode. Start by selecting the desired filters for the analysis, then click 'Next':


Filters selection for Tenant Analysis.
Filters selection for Tenant Analysis.

Then, select the semantic models and workspaces you want to analyze in the "Selection" tab and click on "Run":

Tenant Analysis semantic model selection.
Tenant Analysis semantic model selection.

After authentication, Measure Killer will fetch metadata for the selected items and populate all tabs in Tenant Analysis. The one that provides the information needed in this case is the "Semantic models" tab:


"Semantic models" tab in Tenant Analysis, displaying the different storage modes for all models analyzed
"Semantic models" tab in Tenant Analysis, displaying the different storage modes for all models analyzed

As you can see on the screenshot above, it will show us the various storage modes:

-Import (all data is imported/duplicated in the Power BI semantic model)

-Direct Query (all data is directly queried from the source, like a database)

-Direct query, Import (a mix of other modes -> this is what we call a composite model)

-Direct Lake (a new mode only available in Fabric which is more performant than DQ)


We can now export this list table as an .xlsx file

Exporting the semantic models list/table
Exporting the semantic models list/table

After exporting, I have removed the duplicates in Excel (since we have multiple rows for every model due to the extra information we have in Measure Killer like columns, connected reports etc. in the hierarchy).

My Excel table will now look like the below with the storage mode being in column P (as highlighted on the very right).

Excel output of semantic models with storage mode
Excel output of semantic models with storage mode

8 Ansichten0 Kommentare

Aktuelle Beiträge

Alle ansehen

コメント


bottom of page