Forum Discussion
dataflows: Error native query onprem oracle datasource
I couldn't agree more with Anonymous, not supporting native queries basically limits me to exporting entire tables into the PBI dataset and when handling massive amounts of data plus SQL calcs, this is not what we want.
Hi Anonymous - Are you running into speficic scenarios where query folding isn't producing efficient native queries when you use PQ and M to build a query? Sharing some specific context in the idea (link above) would help the Power BI team understrand the impact of this gap.
Hi MatthewRoche - just to provide a quick example: when selecting a list of active tickets generated within the last 12 months, my database contains tickets generated since 2001. Instead of extracting 450 rows using my Oracle servers as the first layer, I have to extract a list of all tables, then more than a million rows of tickets and then apply M filtering (and a series of filters at that) to achieve the results I want.
And this is oversimplifying the issues. Sometimes my queries will contain joins, subqueries, CASE statements and aggregates that can be processed server-side for ease of implementation, only sending the data I really need for PowerBI.
I will add this to the idea.