Enable Eventhouse endpoint for lakehouse and data warehouse

The Eventhouse endpoint is a powerful capability in Microsoft Fabric that lets you query tables with exceptional speed and ease. Use the Eventhouse endpoint to query lakehouse or warehouse data, discover real-time insights across your data estate, and streamline the analysis of structured, semi-structured, and unstructured data.

Benefits

When you enable the Eventhouse endpoint, you get:

• Instant schema sync: The endpoint syncs tables and schema changes within seconds without manual setup. See the list of sync statuses.
• Mirrored schema: Access current and future lakehouse and warehouse data through a mirrored schema in a dedicated KQL database view.
• Rich consumption and visualization options: Use Copilot, NL2KQL, dashboards, embedded queries, and visual data exploration.
• Reflected in Workspace and OneLake catalog trees: The Eventhouse endpoint and database appear as new branches in your lakehouse tree.
• Fast, scalable queries: Run analytics in KQL or SQL using advanced table operators and commands.
• Advanced insights: Run time series analysis, detect anomalies, and use Python for advanced processing.

After you enable the endpoint, it tracks the source data and optimizes it for Eventhouse performance and flexibility. Each lakehouse or warehouse table attaches to a OneLake shortcut in the Eventhouse endpoint with Query acceleration policies that optimize the source data.

Performance and consumption

On initialization, the endpoint both creates tables and caches data. Within 10 seconds, the endpoint is fully synced. During sync, endpoint query performance slows, but as more data is cached, query performance improves. You can update the cache policy.

The eventhouse endpoint uses a limited portion of the available capacity, which could affect the overall preparation time. While the data is available for queries within seconds, the Query runtime might take longer.

The sync status on the System Overview or Databases page of the Eventhouse helps indicate if cache or sync is in progress. Also check the status for each Shortcut. See sync statuses.

Permissions

Users with contributor or owner permission on the parent data source have contributor permission on the Eventhouse endpoint. They manage cache and retention settings.

Prerequisites

• A workspace in a capacity with Microsoft Fabric enabled.
• A lakehouse or data warehouse with tables in your workspace or in your OneLake catalog.

Enable the Eventhouse endpoint

You can enable the Eventhouse endpoint using any of these three methods:

Method 1: From your Fabric workspace

1. In your Fabric workspace, navigate to your Lakehouse or Warehouse.

2. Select the more options menu (...) and choose Eventhouse endpoint.

   

Method 2: From the OneLake catalog

1. In the OneLake catalog, navigate to your Lakehouse or Warehouse.

2. Select the more options menu (...) and choose Eventhouse endpoint.

   

Method 3: From the Lakehouse/Warehouse ribbon in the OneLake catalog

1. In the OneLake catalog, select your Lakehouse or Warehouse to open the details.

2. From the toolbar, select Analyze Data with > Eventhouse endpoint.

   

Complete the setup

After using any of the above methods:

1. The Eventhouse Endpoint opens with a welcome message. Select Close to start querying the Eventhouse.

   

The workspace and OneLake catalog show the endpoint and KQL database as child items of the Lakehouse.

Query, monitor, and visualize data

The Eventhouse endpoint has these characteristics:

• You can toggle between the Eventhouse, Database, Endpoint, and Queryset pages by using the top bar navigation.

• Queryset: When you create the endpoint, the embedded KQL queryset is named <Lakehouse_Name or Warehouse_Name>_EventhouseEndpoint_queryset. From this queryset, you can Query data in a KQL queryset or use Copilot for writing queries.

  

• System Overview: Shows sync status, storage and activity statistics, and more details including a link to the source lakehouse or warehouse.

• KQL database: The database is named <Lakehouse_Name or Warehouse_Name>_EventhouseEndpoint. As the Eventhouse Endpoint is read-only, create database and add table capabilities are disabled. Run KQL queries, create visualizations in a real-time dashboard, and perform advanced analytics with KQL or SQL.

• Shortcuts: reference OneLake tables that you can query directly using the table function or using natural language via Copilot. If the source has multiple schemas, the schema name shows in each shortcut name. For example, if the schemas are sales and marketing and each has a table named customers, the shortcuts are sales_customers and marketing_customers.

Update the data policy

Update the cache period for a specific shortcut table to optimize query performance. The cache period is measured in days. By default, it's set to 30 days but you can set it to a minimum of one day and a maximum of 36,500 days (approximately 100 years). The availability of data in hot cache increases query performance. See also Eventhouse cost breakdown

1. Browse to your Shortcuts and select the table.

2. From the more options menu ..., or from the top banner, select Data policies.

   

3. In the table data policy section, set the Caching period (days).

   

4. The new cache period applies immediately and is reflected in the Shortcut details pane.

   

Share the Eventhouse endpoint

To share the endpoint, you need to configure share settings for both the endpoint and the source data.

1. Configure permissions to share data in the Lakehouse or share data in the Data Warehouse.

2. Configure permissions to share the Eventhouse endpoint, or to share each KQL database.

3. Send the link to authorized users.

Disable the Eventhouse endpoint

Remove the Eventhouse endpoint from the workspace or the OneLake catalog. Deleting the Eventhouse endpoint also deletes the KQL database, but the Lakehouse remains unchanged.

1. Open your Fabric workspace.

2. Browse to the Eventhouse Endpoint branch of the Lakehouse/Warehouse tree. From the more options menu ..., select Delete.

   

3. In the confirmation dialog box, select Delete.

   If you delete the Eventhouse endpoint, you can re-enable it at any time. The new endpoint creates a new Eventhouse and KQL database, but the new database doesn't retain previous queries, visualizations, or dashboards. If you try to re-enable the endpoint while the previous one is still being deleted, you see a message to wait a few seconds. After the previous endpoint is deleted, you can re-enable it.

   

Reset the Eventhouse endpoint

Use the reset button to delete the endpoint and restart the connection.

Sync statuses

The Eventhouse endpoint syncs source tables and schema changes within seconds. The sync status appears in the System Overview, databases page, and each table shortcut. Here's an example of a sync status:

Sync statuses for the Eventhouse endpoint:

Sync statuses for shortcut:

Considerations and limitations

• You can't enable the Eventhouse endpoint from within an open Lakehouse.
• Changes to the source table schema don't appear at the Eventhouse endpoint.

Related content

• Eventhouse overview
• OneLake shortcuts in a KQL database
• Query acceleration overview