Confluence data source has no connection test or error response in ragflow:latest

[zhanglei012025-zl]

After deploying the service using the ragflow:latest image, I tried to connect a Confluence data source, but there is no button to test whether the connection succeeds or fails, and no error message is returned either.
As a result, I cannot debug whether the issue is caused by the image itself, a configuration problem, or something else.
Is this expected behavior? Is it a known limitation or a bug in the ragflow:latest image?
Also, does RagFlow provide any API endpoint or interface for testing the Confluence connection directly? I would like to verify the connection manually if possible.

[dosubot[bot]]

Hi @zhanglei012025-zl! I'm Dosu and I’m helping the ragflow team.

This is expected behavior in ragflow:latest—there is no dedicated "test connection" button or immediate error response in the UI for Confluence data sources. Connection validation happens internally when credentials are set and during sync initiation, but any errors (like invalid credentials or permissions) only show up indirectly, usually in sync logs or dataset status after you start synchronization. There is no public API endpoint or UI feature for testing the Confluence connection directly; connector creation and update endpoints just save the configuration without verifying the connection at that moment reference.

To manually verify your Confluence connection, make sure your data source is linked to a dataset and trigger a sync. Then check the sync logs and dataset status for errors or missing documents. If nothing appears, common causes include invalid credentials, incorrect wiki base URL, backend service issues, or the sync job not running reference. You can also test your credentials and API access outside RagFlow (for example, using curl to hit the Confluence REST API) to confirm they work.

If you’re running RagFlow in Docker, some images may not support data source sync out of the box—you might need to run the sync script manually inside the container. If you want more detailed troubleshooting steps, let me know!

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[zhanglei012025-zl]

Our Confluence (Server / Data Center with SSO enabled) only allows API access via Authorization: Bearer and does not support authentication using Basic Auth with “username + token/password”.
However, RagFlow’s Confluence connector currently appears to always use Basic Auth for authentication and does not support Bearer tokens. As a result, in environments where SSO is enforced and Basic Auth is disabled, RagFlow cannot connect to Confluence.
Therefore, under the current architecture, the authentication mechanisms of the two systems are incompatible, which makes it impossible to complete data synchronization.
It would be very helpful if the Confluence connector could support Bearer token authentication (for example, by allowing configuration of Authorization: Bearer ), or at least provide a way to configure the authentication type.
In addition, are there any other feasible workarounds or recommended solutions? For example, is it possible to use some kind of proxy to translate authentication, or is there an officially recommended integration approach for environments with SSO + token-based authentication?

[Magicbook1108]

check #11738

[zhanglei012025-zl]

Our Confluence (Server / Data Center with SSO enabled) only allows API access via Authorization: Bearer and does not support authentication using Basic Auth with “username + token/password”.
However, RagFlow’s Confluence connector currently appears to always use Basic Auth for authentication and does not support Bearer tokens. As a result, in environments where SSO is enforced and Basic Auth is disabled, RagFlow cannot connect to Confluence.
Therefore, under the current architecture, the authentication mechanisms of the two systems are incompatible, which makes it impossible to complete data synchronization.
It would be very helpful if the Confluence connector could support Bearer token authentication (for example, by allowing configuration of Authorization: Bearer ), or at least provide a way to configure the authentication type.
In addition, are there any other feasible workarounds or recommended solutions? For example, is it possible to use some kind of proxy to translate authentication, or is there an officially recommended integration approach for environments with SSO + token-based authentication?

[Magicbook1108]

Currently, the Confluence connector only supports authentication via username + token/password.

We are planning to support additional authentication methods for Confluence in the future, similar to what we have already implemented for S3.

If you’re interested, contributions in this area are welcome.

[zhanglei012025-zl]

thanks