Datamart Incremental Refresh vs Refresh

[lg01]

I am trying to understand what the difference is between the two of them. When we do a Schedule refresh, is the Datamart doing a full refresh of every single table? In other words, will it truncate the table and reload all the data from the source? How different is it from the "Incremental Refresh" that can be set up in every table?

[Jonvoge · Accepted solution]

Hi Ig01

Refresh and Incremental Refresh are indeed two different types of refresh.

Both refresh types can be used for Power BI Semantic Models and for Datamarts.

A 'normal' refresh performs a full refresh of the entire dataset. Indeed it is similar to truncating the data and reloading, except if it fails, no data will be deleted, and you will still have whatever was there from the last refresh.

Incremental Refresh requires you to have a date column in the table, and allows you to define periods of time based on that date column, which you want to refresh recurringly, and periods that you want to 'freeze'.

The first time you setup a query with Incremental Refresh, it has to do a Full Refresh similar to a normal refresh. After the initial refresh however, it will only refresh whatever period you defined to be refreshed.

More information about Incremental Refresh in Datamarts can be found here: Get started with datamarts (preview) - Power BI | Microsoft Learn

More information about Incremental Refresh in general can be found here: Incremental refresh for semantic models and real-time data in Power BI - Power BI | Microsoft Learn

_____________________________________________________
I hope my comment was helpful.
If your question was answered, please mark your post as 'Solved' and consider giving me a 'Thumbs Up'.