Query Folding in Teradata

[Anonymous]

Hi,

I am currently connecting to Teradata using ODBC drivers, and notice that some transformations such as row filters, join do not fold in the source database but instead taking place in local memory. Will this feature be supported in future releases of Power BI Desktop or is it because I used ODBC drivers?

[v-yuta-msft]

Hi NickNg278,

"notice that some transformations such as row filters, join do not fold in the source database but instead taking place in local memory"

<--- Could you clarify more details about your requirement?

Regards,

Jimmy Tao

[Anonymous]

For example, if I connect to a table with 100,000 rows, and perfom row filters transformation next (let's say I am filtering sale amount greater than $1000), all 100,000 rows would be first loaded into my computer's memory and filtered out instead of getting only 80,000 rows with sale amount greater than $1000 directly from the source database. 

[v-yuta-msft]

Hi NickNg278,

If you use ODBC connector, you are using import mode other than direct query, so all the raw data are imported from database to your pbix file. You can uncompress your pbix file and check if your metadata are there. When you filter some data, power bi desktop only fetch data from your pbix file, not from database.

Regards,

Jimmy Tao