Premium performance with multiple users on different reports

Hi JoostV,

There is a lot to take into concideration when deciding on a Premium capacity, and then which SKU to choose. I will not be able to cover everything, but I will try to give you some aspects to concider:

1) Storage mode - Import vs. DQ will each have own demands in terms of capacity

Reports in import mode may require extensive memory and CPU during refreshes. You say that there will be no reports larger than 10GB, but long before you reach that size you may already have created one or several reports with long refresh duration. Further, in your example, you use the total of 25 reports. Then you also must consider the refresh schedule, both in terms of parallell refreshes, but also how many refreshes during the 24H.

2) Report interaction

In import mode, when users interact with reports, the model must be uploaded to the memory. If a report is not used in the last 3 minutes it will be regarded as inactive. This allows Power BI to evict the model from the memory when somebody asks to interact with another report that is not already stored in memory. The problem may arise when the memory capacity is reached, and there is no inactive reports to evict. In terms of user experience, at least from my own experience, import models will on average provide a higher level of user experience in terms of waiting time, since all the data is imported and uploaded to the memory.

When it comes to DQ mode, each time someone interacts with the content in the report, Power BI sends a query down to the source. For your Premium capacity, this is demanding in terms of CPU usage, and is positively correlated with the amount of pages and further the amount of visualizations on each page. Also, it becomes even more demanding if RLS is involved.

I am no DBA, so I don't want to say anything about stress on the server.

If you have not already read it, I would suggest you take a look at this article: https://docs.microsoft.com/nb-no/power-bi/service-premium-what-is

Also, in this article you will find a link to another article about the Premium Capacity Metrics App, which might be handy to analyze your environment if you decide to go for Power BI Premium.