When to use Linked tables

[arock-well]

I have six Dataflows in my Power BI Service. All tables in those Dataflows point to various SQL Server tables and these are all refreshing as expected. 

I now want to merge data based on one main table in one of those Dataflows (we’ll call it DF1), but need to look up information in tables from the other various Dataflows (DF2, DF3, etc). I want to merge through the Power BI Service, not in a local Dataset.

Would it be best to first create Linked tables in DF1 to the other Dataflow tables and then merge that way? Or should I look at this in a completely different way? (I’m a newbie so I’m open to suggestions.)

[v-henryk-mstf · Accepted solution]

Hi @arock-well ,

For your needs, if you want to create transformations, merges, etc. for the linked entity, you must create a new calculation entity and refer to the linked entity.

There are several methods as follows: ‎From the data flow authoring tool in Power BI, select ‎‎"Add Linked Entity‎‎"

or select ‎‎" from the ‎‎"Add "Entity‎" menu in the Power BI service to add a link Entity‎‎"

By creating a new data flow entity, you can connect to the data source of the data flow to achieve the effect of merge data from different table.

related document link:

Link entities between dataflows | Microsoft Docs

Connect to data sources for dataflows | Microsoft Docs

Best Regards,
Henry

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[arock-well]

Hi Henry,

Thanks for the great information! What are your thoughts on using a Common Data Model (CDM) in place of or in conjunction with Linked tables and entities?

Aaron