How to Delete Power BI Gateway

Cody Schneider10 min read

Cleaning up your Power BI setup often means getting rid of old, unused components, and the on-premises data gateway is a common one to find on the list. Maybe you’ve moved your data to the cloud or decommissioned an old server. Whatever the reason, this guide will walk you through the full process of deleting a Power BI gateway, from removing it in the cloud service to uninstalling it from your machine.

What is a Power BI Gateway (and Why Delete One)?

In simple terms, an on-premises data gateway acts as a secure bridge. It connects your on-site, private data sources — like a SQL Server database in your office or a massive Excel file on a company server — to a cloud-based service like Power BI. This bridge allows Power BI to reach back into your local network to refresh datasets and keep your reports up-to-date without moving the actual data to the cloud.

So, why would you need to get rid of one? The reasons usually fall into a few common categories:

  • Cloud Migration: The most common reason. Your company has migrated the data source from an on-premises server to a cloud-native one (like an Azure SQL Database or Snowflake). When the data lives in the cloud, you no longer need the on-premises gateway to access it.
  • Decommissioning Hardware: The computer or server hosting the gateway is being retired. The hardware is getting old, and it's time to move the gateway function to a more modern and reliable machine or consolidate it.
  • Consolidation: Your team might have initially set up multiple gateways for different projects. To improve management and reduce overhead, you decide to combine them into a single enterprise gateway cluster for high availability and better load balancing. This means the old, single gateways have to go.
  • Cleanup and Housekeeping: Over time, test gateways or gateways for short-term projects that are now finished can pile up. Removing these keeps your Power BI environment clean and easy to manage, preventing confusion down the line.
  • Switching Gateway Modes: You might need to switch from a gateway running in standard mode to one running in personal mode (or vice versa), which requires a fresh installation.

Before You Delete: A Quick Pre-Flight Checklist

Before you hit the "delete" button, taking a few minutes to prepare can save you hours of headaches and panicked emails from colleagues whose reports have suddenly stopped working. This isn't just about uninstalling software, it's about making sure you don't disrupt your organization's access to data.

1. Identify Gateways and Their Dependencies

First, you need to know exactly which reports, datasets, and dataflows are relying on the gateway you plan to tear down. If you remove it without migrating these assets, their scheduled refreshes will start failing immediately.

Here's how to check dependencies in the Power BI Service:

  1. Log in to app.powerbi.com.
  2. Click the Settings icon (the gear) in the top-right corner.
  3. Select Manage connections and gateways.
  4. Go to the On-premises data gateways tab.
  5. Locate the gateway cluster you're thinking of deleting. You can't see the direct report dependencies here, but you can see a list of connections that use the gateway. This gives you a clear list of the databases, files, and other sources that are relying on that bridge.
  6. Look through the "Connections" and "User" columns. You’ll need to work with those connection owners to ensure they are aware of the pending change.

2. Communicate with Your Users

Data analytics is a team sport. Send out a communication to all report owners and dashboard users who depend on the doomed gateway. Let them know:

  • What is happening: "We are decommissioning gateway 'Old-Finance-Server-GW'..."
  • When it will happen: "...on Friday, October 28th."
  • Why it's happening: "...as we have migrated our finance data to Azure."
  • What they need to do: "Please reconfigure your dataset connections to use the new 'New-Enterprise-Cluster-GW' before this date."

This gives everyone enough time to make the necessary changes and prevents you from becoming the villain who broke everyone's reports on Monday morning.

3. Plan the Migration Path

For every data connection using the old gateway, you need a plan. The plan is usually one of two things:

  • Move to a different gateway: If the on-premises data source is still active, you'll need to point all datasets and dataflows to a different, active gateway. This involves adding the data source connection to the new gateway and then updating the settings for each dataset.
  • Convert to a direct cloud connection: If the data source has been moved to the cloud, you don't need a gateway at all. You can update the data source settings in Power BI to connect directly to the new cloud endpoint.

4. Back Up Your Gateway Recovery Key

While not required for deleting, keeping a record of your gateway’s recovery key is a smart practice. This key allows you to restore or migrate a gateway to a new machine without having to recreate all the data source credentials. Before you begin the deletion process, find your recovery key (you should have saved it during setup) and store it somewhere safe, just in case your plans change last minute.

How to Remove the Gateway from the Power BI Service (Step-by-Step)

Once you’ve completed your checklist and are ready to proceed, the first step is to remove the gateway’s registration from the Power BI cloud service. This severs the connection between the cloud and your on-premises machine.

Follow these steps:

  1. Log into the Power BI Service at app.powerbi.com.
  2. Navigate to the gear icon for Settings in the upper right and click it.
  3. From the dropdown menu, select "Manage connections and gateways."
  4. Ensure you are on the "On-premises data gateways" tab. Here, you'll find a list of all gateway clusters configured in your tenant.
  5. Find the gateway you want to delete from the list. Each entry represents a gateway machine or a cluster of them.
  6. Click the three-dot menu (...) next to the gateway's name to reveal more options.
  7. Select the "Remove" option from the context menu.
  8. Power BI will display a confirmation pop-up. It warns you that removing the gateway cluster will also delete all associated data sources and will cause any reports or dashboards that rely on it to stop refreshing.
  9. Confirm that you understand the consequences and proceed with the deletion.

At this point, the gateway is no longer registered with your Power BI tenant. However, the software is still installed on your server or local machine.

How to Uninstall the Gateway Application from Your Computer (Step-by-Step)

The final step is to remove the actual gateway program from the machine where it was installed. This is a standard software uninstall process done through Windows.

Using the Windows Settings App or Control Panel

This is the most straightforward method for most users.

  1. Log in to the Windows machine (server or PC) that is hosting the gateway application.
  2. If you are using Windows 10 or 11, open the Settings app (you can press Win + I). Click on "Apps." If you are on an older Windows version, open the Control Panel and go to "Programs and Features."
  3. In the list of installed applications, scroll down until you find "On-premises data gateway."
  4. Select the gateway app and click the "Uninstall" button.
  5. A Windows confirmation prompt will appear asking if you want to allow the app to make changes. Click Yes.
  6. The on-premises data gateway uninstaller will now launch. It's a simple wizard that guides you through the removal process. Follow the prompts to complete the uninstallation.

The uninstaller will remove all the files and registry entries associated with the gateway, leaving your machine clean.

Using PowerShell for Advanced Users

If you're managing servers remotely or want to automate the uninstallation, using PowerShell can be much more efficient.

You can identify and uninstall the package with a simple one-line command. Open a PowerShell terminal as an administrator and run the following command:

Get-Package -Name "On-premises data gateway" | Uninstall-Package

This command finds the package with the matching name and pipes it directly to the Uninstall-Package cmdlet, automating the process without any clicks needed.

Common Problems and Quick Fixes

Sometimes, things don’t go as smoothly as planned. Here are a couple of common issues you might encounter during the gateway deletion process and how to resolve them.

Problem: "The 'Remove' option for the gateway is grayed out or I can't see the gateway at all!"

Solution: This is almost always a permissions issue. To remove a gateway from the Power BI service, you must be a designated Gateway Administrator. If you are just a user who can create connections through the gateway, you won't have the privileges to delete it. Check with your Power BI tenant administrator to either grant you the appropriate permissions or have them perform the removal for you.

Problem: "The uninstallation from my Windows machine fails unexpectedly."

Solution: If the uninstaller runs into an error, try these steps:

  1. Stop the Service Manually: Before trying to uninstall again, make sure the gateway's Windows service isn't running. Open the Services app (run services.msc), find the "On-premises data gateway service", right-click it, and select "Stop".
  2. Run as Administrator: If you are opening the Control Panel or Settings app, make sure you are doing so from an administrator account. A lack of permission can stall the process.
  3. Reboot the machine: A simple restart can work wonders. Often, a file needed by the installer is locked by another process. Rebooting releases any such locks and can allow the uninstallation to run cleanly.

Problem: "Oh no! I deleted the gateway before I migrated a critical dataset."

Solution: Don't panic. The report and its data in the Power BI service are not gone - they've just lost their connection to the source data. You cannot "undelete" the gateway configuration. Your only recourse is to fix the data source binding. Open the settings for the broken dataset, and you will see a message that the gateway is unreachable. From there, you can reconfigure it to use one of the existing, active gateways in your environment. You’ll need to re-enter credentials for the data source on the new gateway first, but once you do, the report's scheduled refresh can be re-enabled.

Final Thoughts

Deleting a Power BI Gateway is a two-phase process: first severing the connection in the Power BI Service, and then uninstalling the software from your computer. By following a thorough checklist to manage dependencies and communicate changes, you can ensure the cleanup process is an organized and disruption-free task rather than an unexpected reporting emergency.

Managing on-premises-to-cloud connections can feel old-school and tricky, especially for modern sales and marketing teams who live in cloud-based tools. We designed Graphed to remove exactly that kind of friction. Instead of wrestling with gateways, we connect directly to your most important marketing and sales data sources — like Google Analytics, Shopify, Facebook Ads, and a dozen others — to give you one central place for all your reporting. If you'd rather create powerful, real-time dashboards just by asking questions in plain English instead of debugging gateways, then you might just find our approach a lot simpler.

Related Articles

How to Connect Facebook to Google Data Studio: The Complete Guide for 2026

Connecting Facebook Ads to Google Data Studio (now called Looker Studio) has become essential for digital marketers who want to create comprehensive, visually appealing reports that go beyond the basic analytics provided by Facebook's native Ads Manager. If you're struggling with fragmented reporting across multiple platforms or spending too much time manually exporting data, this guide will show you exactly how to streamline your Facebook advertising analytics.

Appsflyer vs Mixpanel​: Complete 2026 Comparison Guide

The difference between AppsFlyer and Mixpanel isn't just about features—it's about understanding two fundamentally different approaches to data that can make or break your growth strategy. One tracks how users find you, the other reveals what they do once they arrive. Most companies need insights from both worlds, but knowing where to start can save you months of implementation headaches and thousands in wasted budget.