Can anyone explain Quicksight backend query mechanism with spice and without spice

I would like to know and compare, how Quicksight query the data from database for simple TotalSales use case with and without spice

1 Like

Hi @yogendrasgautam - Please find the details below.

  1. When a data set mode is SPICE, QuickSight will fetch the data from database and put the data in SPICE memory. When an analysis is using the data from this data set, it will NOT go to the database and took the data from SPICE which will be very fast. If the same data set is used by many users and visuals, so SPICE is a better option and it will avoid the queries to databases every time. However you need to set up the SPICE refresh depending upon your data freshness. The details on the SPICE can be found from the link - Importing data into SPICE - Amazon QuickSight

  2. When your data set is in Direct Query mode, each request from your analysis is going to run in database directly. If your data is real time it is good to have direct query to show the most updated data as SPICE refresh start at least 15 mins. Direct query is same as like accessing the data from any client tool. So every query will go to database.

Hope the above info provide some details on SPICE and Direct Query.

Regards -Sanjeeb

1 Like

@Sanjeeb2022 thanks,
In case of Direct query, If we create multiple calculated fields or Total Sale, TotalSales by some filter.

Will the quicksight run query on datasource (database) ?
Will it be utilizing the databaase compute resource ?

hi @yogendrasgautam - I believe Yes. To confirm this you can refresh the data set with direct query and calculated field and go to database sessions and you can validate the QuickSight sql as well .

Regards - Sanjeeb

1 Like