topic Re: Embed dashboards in websites and applications in Warehousing & Analytics

Embed dashboards in websites and applications

AE10 — Wed, 13 Nov 2024 14:56:27 GMT

I am embedding a dashboard in my react application using an iframe.

https://www.databricks.com/blog/how-embed-aibi-dashboards-your-websites-and-applications

https://docs.databricks.com/en/dashboards/index.html#embed-a-dashboard

1. How can i let my users view the dashboard, without the need to sign in to databricks?
I want databricks to be transparent for the user so the UX don't feel as if the user navigates away of my app and going through the hole login process.

2. If I have 100 users, for each I want to give access to the same dashboard, only with different dataset, what is the best way to handle that?

Cheers!

Re: Embed dashboards in websites and applications

Alberto_Umana — Wed, 13 Nov 2024 15:31:41 GMT

Hello @AE10

To allow users to view the dashboard without needing to sign in to Databricks, you can publish the dashboard with embedded credentials. This way, all viewers of the published dashboard can run queries using your credentials for data and compute. This allows users to see the dashboard even if they don’t have access to the originating workspace, underlying data, or SQL warehouse. However, be cautious as this might expose data to users who have not been granted direct access to it. https://docs.databricks.com/en/dashboards/index.html#publish-dashboard

To handle access for multiple users, each with a different dataset, you can use the following approach:

Create a base dashboard: Design and publish a base dashboard within Databricks.
Parameterize datasets: Use parameters in your SQL queries to dynamically adjust the dataset based on user input.
Embed the dashboard: Embed the dashboard in your application using iframes.
Pass parameters via URL: Use URL parameters to pass specific dataset identifiers for each user. This way, when a user accesses the dashboard, the appropriate dataset is loaded based on the parameters in the URL.

For example, you can use the following URL structure to pass parameters: <databricks-instance>/dashboardsv3/<dashboard-id>/published?o=<workspace-id>&f_<parameter-id>=<value>

Re: Embed dashboards in websites and applications

bhanu_gautam — Thu, 14 Nov 2024 07:23:36 GMT

@Alberto_Umana , This is really helpful, if we use embedded credentials does it mandatory for end user to have access to data bricks to view dashboard ?

Re: Embed dashboards in websites and applications

AE10 — Mon, 18 Nov 2024 13:49:25 GMT

Hi @Alberto_Umana, in the embed dialog I get this message saying "users will still need to log in..."

Am I missing some settings that I need to configure?

Also, can you maybe provide a code example for the solution you suggested above?

Cheers.

Re: Embed dashboards in websites and applications

Preeti_Singh — Wed, 04 Dec 2024 07:33:05 GMT

Hi ,

Has anyone gotten this working? I am facing the same issue when trying to get the embedded URL working from an iframe in my React application. I think I have all the necessary permissions in place, but still, the Databricks sign-in page opens instead of the dashboard.

Re: Embed dashboards in websites and applications

Zuza-Carvalho — Mon, 21 Jul 2025 16:38:37 GMT

I’ve reviewed the Databricks Embeded Dashboard documentation, which recommends passing parameters via URL encoding when embedding dashboards in external applications or portals.

However, this raises a few concerns from a security and data exposure perspective — especially in multi-tenant environments where dashboards are embedded per customer or user context.

My question is:

What are the best practices or alternative methods to securely pass dynamic filters or parameters to an embedded Databricks SQL dashboard without relying solely on visible URL query strings?

For example:

Is there support for POST-based parameter passing or token-scoped parameters?
Can parameters be derived from a JWT claim or iframe parent context instead of query strings?
Are there supported patterns for hiding or encrypting sensitive filters (e.g., customer_id)?

Thank you !