Single dataset vs individual tenant dataset in multi tenant deployment

Hello QuickSight Community,
We are creating multi-tenant dashboards that will have direct-query datasets connected to the AWS Aurora database. What would be the best architecture for the optimal user experience and system maintenance? If we create a single dashboard for all tenants filtered by tenant-id (one common dataset) then query performance is slow. Another option is to create separate datasets for tenants connected to materialized views then database query performance is optimal but then we will end up with multiple dashboards, datasets, and database objects. Right now, we are not considering SPICE datasets.

Hi @qsuser

Designing multi-tenant dashboards with direct-query datasets connected to AWS Aurora involves finding a balance between query performance, user experience, and system maintenance. Curious to know why SPICE is not considered … Since SPICE helps to get consistent report performance along with scalability & concurrency .

I would strongly appreciate if you could talk to your AWS Account team and ask for design discussion on this usecase . We will be happy to bring in right specialist both from Aurora & QuickSight and recommend the best architecture .

Incase if you don’t have contact details of AWS Account team , please ping me in private message with your contact & organisation details.


Closing this post based on the separate messages. Please let us know if any further help needed on this.