Incremental Refresh, multiple policy(?)

[Remek1]

Hi Guys,

I'm wondering, are there any ways to configure multiple configurations on the SINGLE table?

So for example, I would like to refresh only the last say 1 day on weekdays and a full refresh during weekends?

I Heard this was an option on Analysis Services, but does PBI support something like this through Tabular editor/ DAX Studio or custom M?

Thanks,
Best

[Hariharan_R]

Hi @Remek1 

The default incremental configuration policy supports to create multiple partitions on single table. You can use Tabular editor to create different partitions using M code with different filter conditions.

Ref - https://docs.microsoft.com/en-us/power-bi/connect-data/incremental-refresh-xmla

Thanks

Hari

Did I answer your question? Then please mark my post as the solution.
If I helped you, click on the Thumbs Up to give Kudos.

My Blog :: YouTube Channel :: My Linkedin

[Remek1]

Yes, but my question was more about how, how to schedule incremental refresh with X days during weekdays and full refresh during weekends in an automated fashion, 

Is that even possible?

 

[Anonymous · Accepted solution]

Hi @Remek1 ,

I'm afraid that there is no an elegant way to fulfill your needs. If you set incremental refresh for the dataset, there will be an initial refresh operation when you publish the corresponding report to Service. Partitions must be created, historical data loaded, objects such as relationships and hierarchies are built or rebuilt, and calculated objects are recalculated. But the subsequent refresh operations only refreshes the incremental refresh partition(s), so either individual or scheduled are much faster...

Incremental Refresh in Power BI, Part 1: Implementation in Power BI Desktop

All You Need to Know About the Incremental Refresh in Power BI: Load Changes Only

You can consider splitting the data into historical and current data in two separate tables on the data source side. Then configure schedule refreshes with different date period for these two tables (History table on weekends, current data table on full date period)...

Best Regards