Incremental Refresh: Process data out with Incremental Refresh period

[markhollandau]

Hi,

I have a report set up with an Incremental Refresh looking back 21 days.

Sometimes in my data, a Team Member will need to update a Dimension field that will be out with this 21 day period. How can I keep my Incremental Refresh policy in place while doing a one of refresh to bring back all my data, to capture these changes?

Thanks,

Mark

[markhollandau]

Thanks @lbendlin , that's exactly what I needed.

I can connect SSMS to the Power BI Service Workspace, find the Tables I want to refresh and use SSMS to process the full table, overriding the Incremental Refresh.

Trememdous!

[markhollandau]

Hi @lbendlin ,

How would I force a full refresh? 

Mark

[lbendlin]

By changing the report meta data and re-publishing the report to the workspace.  That triggers a series of events, namely that incremental refresh partitions are deleted and a single partition is created.  After that is complete the next refresh request will recreate the incremental refresh partitions.

Of course this is something you should avoid doing at all cost.  Instead, use the XMLA command for a full refresh of the entire table (see the article I linked to), which will initiate the full refresh of all the table partitions AND of the linked tables/partitions if applicable.  For example if you have auto date/time enabled, this will also refresh all these localcalendar tables.

[lbendlin · Accepted solution]

You can either force a full refresh, or you can refresh individual partitions via XMLA calls.

Advanced incremental refresh and real-time data with the XMLA endpoint in Power BI - Power BI | Micr...

Keep in mind that Incremental Refresh expects immutable data. It is not designed for your scenario (which would require differential refresh)