Edit

Share via

Add an Eventhouse destination to an eventstream

This article shows you how to add an Eventhouse as a destination to an eventstream in Microsoft Fabric event streams.

Prerequisites

  • Access to a workspace in the Fabric capacity license mode (or) the Trial license mode with Contributor or higher permissions.
  • Access to an Eventhouse created in a Fabric workspace where you have Contributor or higher permissions.

Note

The maximum number of sources and destinations for one eventstream is 11.

Add an Eventhouse destination to a default stream

To add an Eventhouse as a destination, you can choose between two ingestion modes: Direct ingestion or Event processing before ingestion.

Direct ingestion mode

Direct ingestion mode ingests your event data directly into the Eventhouse without any processing. You can use direct ingestion mode to add an Eventhouse destination to your default stream or a derived stream.

  1. In Edit mode for your eventstream, select Add destination on the ribbon or select the Transform events or add destination card on the canvas, and then select Eventhouse.

    A screenshot of selecting Eventhouse in the Added destination dropdown list.

  2. On the Eventhouse screen, select Direct ingestion.

  3. Enter a Destination name, a Workspace, and an Eventhouse from the selected workspace.

  4. Select Save.

    A screenshot of the Eventhouse configuration screen.

  5. Connect the new Eventhouse destination card to the output of your eventstream if not already connected, and then select Publish.

    A screenshot of the eventstream with the Publish button highlighted.

  6. In Live view, select Configure in the Eventhouse destination node.

    A screenshot of the published eventstream with the Configure button in the KQL Database destination highlighted.

  7. Your Eventhouse opens in the Get data screen. Select an existing table of the KQL database, or select New table to create a new one to route and ingest the data.

  8. Provide a Data connection name or keep the name provided, and then select Next. It can take a few minutes to pull data from the eventstream.

    A screenshot of the Get data screen for the KQL Database destination, with the Next button highlighted.

  9. On the Inspect the data screen, you can:

    • Select a Format to preview how the data is sent to your Eventhouse.
    • Select Edit columns to configure the columns for your data.
    • Select Advanced to select events to include or to choose mapping options.

    A screenshot showing the data formats and Advanced options on the Inspect the data screen.

  10. If you select Edit columns, on the Edit columns screen you can:

    • Select Add column to add a column.
    • Select Source columns to map.
    • Apply Mapping transformation to columns.
    • Change Sample data values.

    Then select Apply.

    A screenshot of the Edit columns screen.

  11. When you're finished configuring the data, select Finish on the Inspect the data screen.

  12. On the Summary screen, review the details and status of your data ingestion, including the table with the schema you defined and the connection between the eventstream and the Eventhouse. Select Close to finalize the Eventhouse setup.

    A screenshot of the Summary screen with the Close button highlighted.

You can now see the Eventhouse destination on the canvas in Live view.

A screenshot of the configured KQL Database destination in Live view.

Event processing before ingestion

The event processing before ingestion mode processes your event data before ingesting it into the Eventhouse. Use this mode if you apply operators such as filtering or aggregation to process the data before ingestion, or after a derived stream.

  1. In Edit mode for your eventstream, hover over an operator or derived stream, select +, and then select Eventhouse.

    A screenshot of selecting the + symbol for the operator output and selecting KQL Database.

  2. On the Eventhouse screen, Event processing before ingestion should already be selected. Complete the rest of the information about your Eventhouse, and then select Save.

    A screenshot of the KQL Database configuration screen for Event processing before ingestion.

  3. To implement the newly added Eventhouse destination, select Publish.

    A screenshot of the eventstream in Edit mode with the KQL Database destination added.

Once you complete these steps, the eventstream with Eventhouse destination is available for visualization in Live view.

A screenshot of the configured KQL Database event processing flow in Live view.

Note

When configuring an Eventstream, the source, transformation logic, and destination are typically added together. By default, when publishing the Eventstream, the backend services for both data ingestion and data routing start with Now respectively. However, data ingestion might begin faster than data routing, causing some data to be ingested into Eventstream before routing is fully initialized. As a result, this data might not be routed to the destination.

A common example is a database Change Data Capture (CDC) source, where some initial snapshot data could remain in Eventstream without being routed to the destination.

To mitigate this, follow these steps:

  1. When configuring an Eventhouse (Event processing before ingestion) or Lakehouse destination, uncheck Activate ingestion after adding the data source.

    A screenshot of the KQL Database without selecting Activate ingesting after adding the data source.

  2. Manually activate ingestion after the Eventstream is published.

  3. Use the Custom time option to select an earlier timestamp, ensuring initial data is properly processed and routed.

    A screenshot of resuming the KQL Database. For more information, see Pause, and resume data streams

Related content

To learn how to add other destinations to an eventstream, see the following article: Route events to destinations