A Matter of Size

When you work with Power Bi Report Server, one of the things you must always monitor is the actual RAM occupation by published reports.
This one does not correspond to the size of the .pbix files you have published and therefore the size that you can check by accessing the properties of the published report does not correspond to the information you are looking for.

Indeed, when we update a published report with the new data available, we would expect the model size to increase, but the report size does not change.
Why?
Because what we see is the size of the report itself, not the data model underlying the report.
Everyone who works with Power BI knows that the data model we create when we build a report is nothing more than an instance of Analysis Services. So if we want to find out the RAM occupied by our report, we need to find our model and find out its size.
But where is this model?
It is located on an Analysis Services instance that it is created when we install Power BI Report Server. We can access it connecting with Management Studio to the same server indicating port 5132.

In this instance we will find a database for each Power BI report published on Reporting Services.

Each database has as its name, the ItemID with which the corresponding report was saved in the Catalog table of the Report Server database.

Once we have found the database we are looking for, by clicking on the properties we can easily discover the actual size of our model.

If we want to know more about the size of our model, we can use a very powerful tool like DAX Studio.

If you don’t know DAX Studio, you can download it at https://daxstudio.org/
We can open DAX Studio and connect to the desired instance.

In the Advanced section we find the View Metrics function.

By clicking on View Metrics we activate on our database Vertipaq Analyzer, a very useful tool that provides us with a lot of information about the structure, the goodness and the issues of our model.
In our case the first thing we can check is the overall size of our model in the Summary section.

Later we can investigate the detailed information to find out which tables and which columns weigh the most in our model and possibly understand how to improve the performance of our model.

But this is a vast topic that deserves (at least) a dedicated post.
Obviously you can use DAX Studio to analyze your model connecting directly to your .pbix file before you upload it on the report server or on the Power BI Service.