Difference between writing a custom SQL on Quicksight when your data source is Athena vs Creating a view on Athena itself and then using it on Quicksight

Hi People,

Q1 Need to understand a difference between writing a custom SQL on Quicksight when your data source is Athena vs Creating a view on Athena itself and consuming it in Quicksight?

Q2 If I set Incremental load on above both of the situation, Quicksight will scan entire Database to load incrementally, right?

@shweta ,

For Q1 , i have explained it here : Incremental Refresh Limitations Question - #9 by Koushik_Muthanna
For Q2 , I believe yes , though you would have to test it out when the date is field is used in partitioning.

Kind regards,
Koushik

Hi @shweta - If you are creating the view in athena with the custom sql, there is NO difference on the execution time and result. at end of the day the same set of sqls are going to be executed in Athena. However if the custom sql is not changing a lot and kind of stable, i will suggest to create the view as it will help you in the operational prospective and you can avoid users to update the logic from UI side.

On the incremental approach, it depends how you are calculating your incremental data but both the cases the result should be same.

Regards - Sanjeeb

1 Like