Dataflow Gen2 and nested tables

[jsconcepcion]

Im working on a Power BI report development project and the main data source is a datalake which I connect via ODATA. I have setup few dataflows for it. Ive read the documentation and found out that the best practice is to separate the staging from the transformation layers. 

So I created a staging one where I load some tables from the datalake without any steps/transformations after. Then I created a second dataflow and point to the staging one as a reference. Now from here, I noticed that all columns that contains nested tables are gone. After I publish the dataflow and come back later, an additional step was automatically added that removes all columns with nested tables. I really need those columns

How can I implement the best practice if this is the case? Do I need to load absolutelty all tables from the datalake for the nested columns to not disappear?  There are about 1000+ tables and we dont need all those data, only a handful.

Thanks for reading,

Justine

[lbendlin · Accepted solution]

After I publish the dataflow and come back later, an additional step was automatically added that removes all columns with nested tables. I really need those columns

in order to be used in a dataflow or dataset semantic model the result of a Power Query must be in tabular format and must contain only basic column types.  No objects (like nested tables) are permitted.

[jsconcepcion]

I can see that now. Thank you!