Forum Discussion
Power BI Report Server Scale-Out Issue - Reports Visible from One Node but Empty Folder from Another
Post PartisanHi sthirumala
I’ve seen this exact behavior before with PBIRS in scale-out + AG setup.
99% this is an Encryption Key mismatch between the 2 nodes.
When Node A shows "This folder is empty" but Node B shows everything, it means Node A can't decrypt the Catalog/Content from the shared ReportServer DB.
Here’s what to check
1. Sync Encryption Keys : Go to PBIRS Config Manager on Node A → `Encryption Keys` → `Restore` → Use the key file exported from Node B.
Scale-out nodes MUST have identical encryption keys.
2. Check Service Account : Both PBIRS services should run under the same domain service account with proper permissions on the AG Listener.
3. Clear Cache : After restoring keys, restart both PBIRS services and clear browser cache.
4. Verify Web Portal URL : Make sure both nodes have the same Web Portal URL and root folder permissions.
Try step 1 first. Restoring the key from Node B to Node A usually fixes it immediately.
Let us know if that works!
Thank you for the reply.
I have restored the encryption key from Node B to Node A; however, the issue still persists.
The behavior is quite unusual. Regardless of which node is hosting the primary AG replica (Node A or Node B), accessing the web portal through the cluster name over HTTPS consistently displays a "Folder Empty" message on the primary replica.
However, when accessing the web portal directly using the individual node name (HTTP on port 80), the folder contents are displayed correctly from both nodes without any issues. This suggests that the report content itself is accessible and functioning as expected.
The issue appears to be specific to accessing the web portal via HTTPS using the cluster name object.
As part of the troubleshooting process, I have already:
- Restored the encryption key from Node B to Node A
- Verified that the same service account is being used on both nodes
- Cleared the browser cache/history
- Reviewed the Report Server logs
Despite these steps, the behavior remains unchanged, and I do not see any related errors or warnings in the Report Server logs.
Additionally, we have a similar setup in our non-production environment that is working as expected without any issues, which makes this behavior even more puzzling.
Thank you for your assistance.
- v-abhinavmu
Community Support
Hi sthirumala,
Thank you for sharing the additional troubleshooting details. and thanks to Gautam_Kumar01 for sharing valuable insights.
Based on the Microsoft documentation for Reporting Services, the following configuration requirements apply to a scale-out deployment:
• Reporting Services creates and manages the URL reservations for both the Report Server Web Service and the Web Portal through Reporting Services Configuration Manager. The tool creates the URL reservations, assigns the required permissions, checks for duplicate instances, and registers the URLs with HTTP.SYS. Microsoft also recommends not creating or modifying Reporting Services URL reservations manually using other tools.
• For a scale-out deployment that uses a network load-balanced (NLB) cluster, Microsoft states that the NLB virtual server name should be used in the configuration of the report server URLs.
• Microsoft also states that the report server instances in a scale-out deployment should share the same ReportServer database.
Could you please confirm whether your deployment is configured in accordance with these documented requirements, particularly with respect to the report server URL configuration for the cluster/NLB virtual server name?
For more details, please refer to the below official documentation:
URL reservations & registration (Report Server Configuration Manager) - SQL Server Reporting Services (SSRS) | Microsoft LearnI hope this helps. Please feel free to reach out if you have any further questions.
Thank you.Hi v-abhinavmu
Thank you for your reply. I confirm that,
- Report web URLs were configured using reporting services configuration manger only. Both the nodes use the same URLs with clustername in it.
- We don't have any network load balancer in place.
- Both the nodes use the same report server databases using AG listener.
I would like to reiterate the same setup works fine in non prod build. But, it is weird with the new production build.
Report web portal displays folders either from secondary replica or client laptops. But, regardless of which nodes is primary replicas, it always shows folder is empty message.
So, the moment we access web portal locally from primary replica, it hows "This folder is empty."
I have tried reinstaling powerbi report server and reconfiguring multiple times. But no luck.
