Forum Discussion
Cloud connection and Data source credentials
Hi Community,
We have been using on-premise gateway all the time for the semantic models that connect the Azure Databricks. Lately, we have been playing around with the Databricks new feature: Power BI task flow and we'd like to remove the data gateway in between.
Since I'm connecting to cloud data sources, I have Cloud connections and Data Source Credentials. Both seem to work if I use a personal connection under the Cloud connections section or if I put the credentials under the Data source credentials.
What is the difference between those two? The semantic models will be used across the organization, what would be the recommeded way to set it up?
In my previous experience dealing with on-premise gateway, I have scripts to call the REST API to bind the gateway for any new published semantic models. Will there be any REST APIs that I can utilize to automate the binding process for either cloud connection or authorizing the Data source credentials?
Appreciate any comments!
Hi dmkblesser ,
This is something a lot of us are navigating with Databricks and the move away from on-prem gateways.
Here’s the bottom line:
-
For anything org-wide or production, always use Shareable Cloud Connections. They’re designed for centralized, secure, and reusable credential management. Setting up at this level means fewer headaches down the road.
-
Individual Data Source Credentials are really only for quick testing or personal stuff. They don’t scale and can lead to messy governance if used for shared models.
-
For automation: Right now, there’s no REST API to directly bind a semantic model to a Shareable Cloud Connection (unlike gateway binding). You’ll need to assign the connection manually for now but keep an eye on Microsoft’s roadmap, because I bet this will change soon.
-
4 Replies
Hi dmkblesser ,
Prioritize Shareable Cloud Connections: For organization-wide use and better management.
Databricks Power BI Task Flow: Great move to eliminate the on-prem gateway. Ensure your Databricks connection in Unity Catalog is set up appropriately for orchestration.
Automation: While the Gateways - Update Datasource API exists for credential updates, a direct, comprehensive API for programmatically binding newly published semantic models to existing shareable cloud connections might still have some gaps or require specific workarounds. Keep an eye on the Power BI REST API documentation and community forums for updates on this specific automation.
Hope this helps you navigate the Power BI cloud connection landscape!
Best Regards,
Jainesh Poojara | Power BI DeveloperThankyou, jaineshp, for your response.
Hi dmkblesser,We appreciate your inquiry submitted through the Microsoft Fabric Community Forum.
Based on my understanding, the confusion appears to stem from overlapping methods of credential management. Specifically, Cloud Connections, particularly Shareable Cloud Connections, are best suited for enterprise usage, whereas Data Source Credentials apply to individual datasets and therefore lack scalability for organization wide implementation.
If both methods are configured simultaneously, Power BI may permit either option to function, but this often results in inconsistent credential governance and increased complexity in automation.
For usage across the entire organization, it is recommended to utilize Shareable Cloud Connections to ensure consistency, reusability, and simplified management of credentials. The use of individual Data Source Credentials should be avoided except in testing or personal use scenarios.
At present, REST APIs do not support the direct binding of semantic models to Shareable Cloud Connections. While APIs such as ‘Update Datasources’ are applicable for gateway based sources, they are not fully compatible with cloud connections.
We hope that the information provided will assist in resolving your query. Should you have any further questions, please do not hesitate to contact the Microsoft Fabric Community.
Thank you.
Hi dmkblesser ,
This is something a lot of us are navigating with Databricks and the move away from on-prem gateways.
Here’s the bottom line:
-
For anything org-wide or production, always use Shareable Cloud Connections. They’re designed for centralized, secure, and reusable credential management. Setting up at this level means fewer headaches down the road.
-
Individual Data Source Credentials are really only for quick testing or personal stuff. They don’t scale and can lead to messy governance if used for shared models.
-
For automation: Right now, there’s no REST API to directly bind a semantic model to a Shareable Cloud Connection (unlike gateway binding). You’ll need to assign the connection manually for now but keep an eye on Microsoft’s roadmap, because I bet this will change soon.
Sharing the conversation with a Microsoft solution engineer:
The Power BI REST APIs do not support binding datasets to Shareable Cloud Connections. While Fabric has introduced a new update-connection API, it's still limited and doesn't fully replace gateway binding for cloud sources. For now, cloud connections must be configured manually. I have no visibility into the roadmap beyond what we make publicly available here. This is a known gap, please vote up related ideas.
This is the idea I found that related to this topic and I voted. https://community.fabric.microsoft.com/t5/Fabric-Ideas/REST-API-to-Update-Gateway-Cloud-Connection-for-Power-BI/idi-p/4738079
-
