Power bi load testing on the services

[jnjAnil]

I want to test what happens if 50 concurrent users access my Power BI dashboard on the service. Currently, I’m using F64 capacity with an import mode data model. I’d like to check whether this setup is sufficient or exceeds the expected capacity

[GilbertQ]

Hi @jnjAnil 

You can watch this video below of myself where I explain and show you how to use the steps highlight above but in a video format. You can follow along and get it working as expected. Power BI Performance Testing Tool by Gilbert Quevauvilliers

[anilgavhane · Accepted solution]

To test how your Power BI dashboard performs under 50 concurrent users using PowerShell and Git, you’ll need to simulate load and monitor performance. Here's a step-by-step guide to help you set it up:

Prerequisites

1. Power BI Load Testing Tool
   Microsoft provides a Power BI load testing tool on GitHub that uses PowerShell to simulate concurrent users.
2. Git Installed
   So you can clone the repository.
3. PowerShell 5.0+
   Required to run the scripts.
4. Power BI Premium Capacity (F64)
   Ensure your dashboard is published to a workspace backed by F64 capacity.

Step-by-Step Setup

1. Clone the GitHub Repository

git clone https://github.com/microsoft/powerbi-load-generator.git cd powerbi-load-generator

2. Install Required Modules

Install-Module -Name MicrosoftPowerBIMgmt -Scope CurrentUser

3. Authenticate to Power BI

Login-PowerBI

4. Configure the Test Parameters

Edit the config.json file in the repo to define:

• Report ID
• Workspace ID
• Number of users (set to 50)
• Duration
• Refresh interval

Example snippet:

{ "reportId": "your-report-id", "workspaceId": "your-workspace-id", "numberOfUsers": 50, "durationInMinutes": 10, "refreshIntervalInSeconds": 30 }

5. Run the Load Test

.\Start-LoadTest.ps1 -ConfigFilePath ".\config.json"

This will simulate 50 users accessing the report and log performance metrics.

What to Monitor

• CPU and memory usage in the Power BI Premium Capacity Metrics app
• Query duration
• Visual load time
• Throttling or failures