ISSUE: Fabric SQL database replication error to one-lake

[Max_agileDSS]

Hi everyone,

I've been facing a weird error since few days.

I created a SQL Database in a workspace (Canada Central), but the mirroring replication process to OneLake doesn’t seem to be working and is showing an error.

here is the error: 

Code: SqlError, Type: UserError, Message: The server principal "blabla@blabla.com" is not able to access the database "sql_DataVex_Metadata-123456789" under the current security context. ArtifactId: 123456789.

I also asked a colleague to test it in another project/tenant, and they encountered the same issue.

Furthermore, I can confirm that this feature was working fine last week, since I have another workspace with a SQL Database where everything is currently working as expected.

So my questions are:

• Is anyone else experiencing the same issue/behavior ?

• Has Fabric changed this feature since the last release ?

• Is there a new configuration to perform ?

Thanks in advance for your help !

[Poojara_D12]

Hi @Max_agileDSS 

The error you are seeing with SQL Database mirroring to OneLake is not caused by your setup alone but rather by how Microsoft Fabric now enforces security around mirrored databases. The message “The server principal is not able to access the database under the current security context” usually means the identity that Fabric uses to replicate data—specifically the System Assigned Managed Identity (SAMI) of the Azure SQL server—does not have the right access. Since you mention this was working fine last week and now fails in multiple tenants, it is likely due to a recent change in Fabric or Azure SQL permissions. To fix it, you should verify that the SQL logical server’s SAMI is enabled, that it is the primary identity (and not overridden by a User Assigned Managed Identity), and that it has Read and Write permissions on the mirrored database artifact in Fabric. You should also confirm that your workspace or tenant allows OneLake external access, and that the SAMI has been granted access to the source SQL database itself. In short, the feature itself hasn’t disappeared, but security enforcement appears to have tightened, so ensuring the SAMI has correct rights and configuration should resolve the replication error.

[v-pnaroju-msft · Accepted solution]

Thankyou, @ToddChitt and @j_hoff for your responses.

Hi Max_agileDSS,

We appreciate your inquiry submitted through the Microsoft Fabric Community Forum.

The backend team has resolved the issue, as shown in the screenshot below:

Please confirm whether you are still experiencing the problem. If you have any further queries, feel free to contact the Microsoft Fabric Community.

Thank you.

[Max_agileDSS]

Hi @v-pnaroju-msft ,

I can confirm that the applied fix has resolved the issue as expected.

I’ve checked all existing SQL Databases and also created a new one to be sure and replication worked for all of them.

Thank you very much and have a nice week ! 🙂

p.s. And thanks @j_hoff  + @ToddChitt  for helped me to find a workaround while waiting for the fix.

[ToddChitt]

My client issue is still not resolved. I am working through a Microsoft Support Ticket (opened last Friday) and hope to get some answers soon.

[Max_agileDSS]

Thanks @j_hoff  and @ToddChitt !

I've checked the note on the service status.  However, this doesn’t seem to be exactly the same issue, but it might be a side effect… I’ll keep a close eye on it.

I’ll also ask my team to submit an issue to make sure they are working on it, since I’m still facing the problem.

I’ll let you know if I notice any changes.

Thanks !

[j_hoff]

I'm not sure if it was something I did or if the issue was resolved by Microsoft, but my Fabric SQL Database is replicating as intended again.

About 20 minutes ago, in the Fabric SQL Database I went to the Security tab -> Manage SQL Security, and added myself to all of the four built-in security roles. Did this change anything? I have no idea, as I am the owner of the SQL Database item. After doing this, I tried to "bump" the replication process by adding a column to my table that was showing a status of "RunningWithWarnings". I verified the new column showed up in the table and then dropped the column. I checked back a few minutes later and my table is now showing a status of "Running" in the Monitor Replication tab. I checked that the two columns I had tried to originally create had replicated to the SQL Analytics Endpoint and OneLake, and they had. I also created a new table in the dbo schema and it also replicated as intended.

I'm curious what you all find!

[Max_agileDSS]

@j_hoff   This is a good lead and I think i've found a "dirty workaround"...

Here is the sequence that I did:

• Created a SQL database.

• Explicitly added the item owner to all four database roles.

• Important: Even after adding the owner to the roles, I still got an error. So I paused and resumed the Fabric Capacity to make sure that i've forced db's to restart...

• Waited until capacity is running, then execute EXEC sp_help_change_feed to check that the change feed/replication process is working ... 

• Added tables and data and the replication process ran successfully.

Although this seems to work, it remains a temporary solution that I can apply individually. It is not viable for my customer because there are users assigned to the current capacity and haves multiple SQL Databases. Nevertheless, it can be used as a desperate workaround. 🙂

Can you confirm that this sequence seems to work?!

Thanks! 🙂

[j_hoff]

That logic makes sense to me, I don't have a seperate capacity that I could try it on though.

In other news, it appears we've made it on to the support page 🙂

[j_hoff]

I'm experiencing the exact same issue. The replication had been working fine for several months until I altered a table (added two new columns) a couple of days ago. I noticed that the two new columns were not populating in the SQL Analytics Endpoint so I checked the replication status and it is showing "RunningWithWarnings" for the table I had altered. I then created a brand new table in the dbo schema. The replication status for the brand new table shows "Failed". I then inserted a record into an existing table and it replicated to OneLake within a minute. Running the DMV's provided in Microsoft's troublshooting guide also did not return anything out of the ordinary. I saw that there is a Service Status issue reported on Microsoft Fabric's support site that involves mirroring and a fix is expected to be deployed by end-of-day 9/13/2025. I do not know if the two are related, but I've got my fingers crossed that it will get fixed.

[ToddChitt]

Can you share the link to that issue/post on Fabric Status?

[j_hoff]

Microsoft Fabric service status

[ToddChitt]

@Max_agileDSS 

YES! I am seeing the same thing. For one of our clients that has a SQL Database in Fabric, it started last Sunday, or as early as 9/4/2025. I have created a test case in our environment by creating a table in a Fabric SQL Database, shoving in about 2,000 rows, then seeing how long it takes to get synchronized over to OneLake and the SQL Analytic Endpoint. After TWO HOURS it does not even see the table yet, let alone any rows.

I call BUG!