Balancing Memory and Database Usage on Load

Hi sem1 ,

For your question 1:

I'm afraid that there is not option for us to stop the native query in power query. Actually, we only could view the native query in the applied steps in query editor.

For your question 2:

Yes, when developing the model, this in-memory load will be on your PC. And when refreshing via the data gateway, the in-memory load will be the server which install the gateway to read the source, but the in-memory also load in cloud in could when deal with the data source in Azure service bus.

You could have a good look at the document to understand it better.

Best  Regards,

Cherry

---

v-piga-msft wrote:

Hi sem1 ,

For your question 1:

I'm afraid that there is not option for us to stop the native query in power query. Actually, we only could view the native query in the applied steps in query editor.

 

---

Can't you use a transformation not supported by native queries and that will force the remainder of the transformations to occur via power query in-memory?

Edit; Tested this and yes I can. How I'm doing it is left joining tables from the data source to a table I input with one non-matching key then deleting the column after expanding it. (This leaves the data source the same as it was before the join). Because cross-data source merges can't be done in Native Query, Power BI then loads the data into memory and processes it for all further steps.