Microsoft Fabric Updates Blog

Updates to default data destination behavior in Dataflow Gen2

When you have a Lakehouse or Warehouse and you want to load data into it, you can use Dataflow Gen2 as an easy, low-code way for landing your data with the right shape.

You can always create a stand-alone Dataflow Gen2 and use the data destinations to load your data in any Fabric Lakehouse or Warehouse, but to speed up your development, there are some easy other ways to land your data faster.

This blog is to update you on the experience and important changes that were made.

Within the Lakehouse or Warehouse experience, you can get data through a variation of options.

Select new dataflow gen 2 from get data dropdown in Lakehouse or Warehouse

When choosing Dataflow Gen2 from either the Lakehouse or the Warehouse, the data destination experience is slightly different from a ‘standard’ Dataflow Gen2 that was created from the workspace.

By default, any query that you create will have the Lakehouse or Warehouse you started from set as the data destination. If you hover over the data destination icon, you can see that the destination is labeled as ‘default destination’. This is different from the standard Dataflow Gen2, where you explicitly have to assign a query with a data destination.

Hover over the data destination icon to see that the destination is labeled as defautl destination

With the default destination, the settings are set to a default behavior that cannot be changed. This are the behaviors for both Lakehouse and Warehouse default destination:

BehaviorLakehouseWarehouse
Update methodReplaceAppend
Schema change on publishDynamicFixed

Note: Previously, update method for Lakehouse was append. This is now changed to replace.

To edit the settings of an individual data destination, use the gear icon to edit the destination. When you edit the individual data destination, this change is only affecting the specific query. It is currently not possible to change the behavior of the default destination.

Related blog posts

Updates to default data destination behavior in Dataflow Gen2

April 16, 2024 by Ruixin Xu

We are pleased to share a set of key updates regarding the Copilot in Microsoft Fabric experiences. The information in this blog post has also been shared with Fabric tenant administrators. Below are the highlights of the changes. This change is an important milestone to eventually allow Copilot to reach GA within this year. See … Continue reading “Copilot in MS Fabric: Soon available to more users in your organization”

April 15, 2024 by Santhosh Kumar Ravindran

Users orchestrate their data engineering or data science processes using notebooks and in most of the enterprise scenarios pipelines and job schedulers are used as a primary option to schedule and trigger these Spark jobs. We are thrilled to announce a new feature Job Queueing for Notebook Jobs in Microsoft Fabric. This feature aims to … Continue reading “Introducing Job Queueing for Notebook in Microsoft Fabric”