How to update all Fact and DIM tables in Fabric at the same moment

[Kazanskyi]

I am working on moving the current Power BI dataset to the Fabric.

The data source is an on-premise SQL data warehouse that refreshes multiple times per day, and refresh is usually over during working hours.

I am copying tables using a pipeline from the warehouse to Fabric, and Fact_Tables and DIM tables take different amounts of time to load.

As far as I understand, Power BI reads from the latest version of the table in Fabric, and data is available immediately after the table is updated.

This means that users may get the wrong data at some point during the day because the DIM and FACT tables are not updated synchronically.

How to achieve the same full-package delivery as the Power BI dataset?

[tackytechtom · Accepted solution]

Hi @Kazanskyi ,

I happened to have written a blog article about this exact scenario. Feel free to check it out:

https://www.tackytech.blog/how-to-refresh-your-direct-lake-semantic-model-from-fabric-data-pipelines...

In this blog post, we disable the setting which keeps the Direct lake data up to date. After, we trigger a "refresh" (or in this case it's rather referred to as reframing) from a fabric data pipeline. There is even a new dedicated pipeline activity for this matter, too.

Let me know if this helps you or in case you have questions along the way 🙂

/Tom
https://www.tackytech.blog/
https://www.instagram.com/tackytechtom/