Power BI Premium capacities give enterprises access to their own dedicated resources in the cloud. With this access enterprises can distribute Power BI content as widely as they choose; it also lets them unlock greater scale, better performance and advanced capabilities that are not available to Power BI users on shared capacities. All of these are possible because you can choose how a Premium capacity’s resources get utilized. 

Today we highlight a recently released tool that can help you understand how much user load your capacity can handle when used. During Microsoft Business Application Summit in Atlanta last June, we introduced a tool for automating load testing  of a Premium capacity. A video run through of the tool is available here ( watch from the 31:00-minute mark)

The tool uses PowerShell to create automated load tests against your capacities, and lets you choose which reports to test and how many concurrent users to simulate.

The tool will generate load on a capacity by continuously rendering each report with new filter values (to prevent unrealistically good performance due to report caching) until the token required for authenticating the tool against the service expires. 

Using the new tool, capacity administrators can get a better understanding of how many users their capacity can handle in a given time frame.

The tool can also be used by report authors who want to understand the user load effect of performance improvements measured with Power BI desktop’s Performance Analyzer 

While the tool is running, users are able to both see the renders happening in real time on their browsers, as well as connect to the XMLA endpoints of the capacities being measured with SQL Server Profiler to see the queries execute as they come.

As the load test is executing, its effects will also be visible in the premium capacity metrics app’s Datasets page. Capacity admins can use this tool to generate load and see how that load shows up. Be sure to read the Premium Deployment whitepaper section about identifying overload in the metrics app beforehand.

To see the effects of the load test in the metrics app after the test has been run, admins should expand the “datasets” tab of their metrics app landing page and initiate an on-demand refresh by clicking “refresh now”. Admins should expect up to a 15 minutes lag from the time the test starts generating load until the time that load is visible in the metrics

To host the tool and other future tools and utilities,  we established a new GitHub repository. The entire repository is Open Source and users are welcomed to contribute, add other tools and repositories related to Power BI Premium and Embedded capacities and improve the existing ones. 

Please be mindful of the existing load on your capacities and make sure not to run load tests during top usage times.
We look forward to hearing your feedback from using the tool and your contributions to the repository!

The Power BI Premium team

Let’s block ads! (Why?)

Microsoft Power BI Blog | Microsoft Power BI