Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Find everything you need to get certified on Fabric—skills challenges, live sessions, exam prep, role guidance, and more. Get started

Reply
Anonymous
Not applicable

Only refresh changed rows possible with Scheduled Refresh

Hi Community,

 

Need some help once more since I'm struggeling at one of my projects.

The situation is as follows. We have a big amount of sales and purchase order lines combined in to 1 table. All data we get from SQL Database. We currently do not have a DateTimeLastModified field on the rows but this could be added if it is a solution.

We use Import and have 8 scheduled refreshes during working hours. 

This takes up alot of resources however with only a handful of rows changing between each refresh. I've looked into the possibility of using scheduled refresh functionality but I'm not 100% sure that this works as a solution in our case. Since all the rows in our dataset are subject to changes and never 'final'.  It seems that with scheduled refresh you only set a timeframe in which records should be refreshed. So if I set RangeStard to 9-4-2021 and RangeEnd to 10-4-2021, only these records are refreshed:

 

Koense_0-1628771011310.png

 

Or could we just add the DateTimeLastModified to the rows, then set up RangeStart as minimum date and RangeEnd as maximum date and set the Detect changes field as the DateTimeLastModified?

Koense_1-1628771186766.png

 

Koense_2-1628771227051.png

 

Please let me know what the best way to go is. 

1 REPLY 1
lbendlin
Super User
Super User

Incremental refresh is not entirely suited for what you are trying to achieve.  What you are looking for is differential refresh (what Qlik calls CDC).

 

Absolutely DO add the DateLastModified column to your data, it is extremely useful.

 

BUT!  You cannot specify the same field for RangeStart/RangeEnd  and for Change detection.  These have to be fed from separate fields.

 

What we ended up doing was to use a hybrid approach where we would fetch the data in reasonable chunks (for example fiscal quarter, or "transactions created in the last three months"  , with the assumption that the data would have calmed down after that period,  and then flush and fill only these chunks as needed. 

 

We do use incremental refresh, but our RangeStart/RangeEnd is based on the chunk dates, not the data dates  (we don't have the luxury of the DateLastModified field in our data)

Helpful resources

Announcements
Europe Fabric Conference

Europe’s largest Microsoft Fabric Community Conference

Join the community in Stockholm for expert Microsoft Fabric learning including a very exciting keynote from Arun Ulag, Corporate Vice President, Azure Data.

AugPowerBI_Carousel

Power BI Monthly Update - August 2024

Check out the August 2024 Power BI update to learn about new features.

August Carousel

Fabric Community Update - August 2024

Find out what's new and trending in the Fabric Community.

Top Solution Authors