Fails to Sync SQL database to SQL endpoint

[debiagui]

We’ve been working with a SQL database for the last few months. Today, as usual, we inserted new rows into a table, but even after waiting 5 hours, those rows are still not appearing in the SQL endpoint.

We also tried triggering the SQL endpoint sync using the REST API, but there are still no changes.

Any recomendation? link to read? or know issue?

[debiagui · Accepted solution]

My issue is related to this another question in the community

ISSUE: Fabric SQL database replication error to on... - Microsoft Fabric Community

I'll wait for the fix

[debiagui]

Hi @Poojara_D12 yes please can you guide me to the lasted know issues?

I created a new dabase and replicate all data and again fails to update sql endpoint, seems like is a know issue.

My tables are normal, no custom column, or any extrange index logic  

[v-sgandrathi]

Hi @debiagui,

We recognize your concern regarding the SQL endpoint not updating even after creating a new database and replicating the data. Currently, this is a known limitation, as synchronization between Lakehouse data and the SQL endpoint may take longer than anticipated. Factors such as the number of tables, partitioning structure, and delta log files within Lakehouse can contribute to these delays.

To address this, users have found that waiting approximately 10–15 minutes after loading data before querying the SQL endpoint often ensures that new rows are visible.
Additionally, Microsoft has introduced a documented “Refresh SQL Endpoint Metadata” REST API, which can be called programmatically post-ingestion to trigger synchronization and monitor its completion status, thereby improving reliability.

We also suggest implementing best practices like vacuuming or checkpointing delta tables, especially in environments with substantial partitions or large logs, as these steps can help maintain consistent synchronization times.

Thank you.

[debiagui]

Hi @v-sgandrathi the issue is with the data in the SQL database versus the data in the SQL endpoint of the same database. I created a new database and replicated the schema without any indexes, relationships, or special columns—just varchar, integer, and long fields—but the SQL endpoint does not show any tables.

[v-sgandrathi]

Hi @debiagui,

Thank you for providing the additional information. If the SQL endpoint is not displaying any tables, even after creating a new database with a basic schema, this likely indicates a metadata registration issue rather than just a problem with row synchronization.

To address this, please consider the following steps:

• Confirm that your tables are registered in the Lakehouse view under the Tables section. If they appear only as files and are not promoted to tables, they will not be visible in the SQL endpoint.

• Use the Refresh SQL Endpoint Metadata REST API to manually refresh the schema. This can help in cases where new tables are not automatically recognized by the endpoint.

• Verify that your database is created within a Fabric-enabled workspace and that the SQL endpoint feature is active. Workspaces without Fabric support will not expose schema in the SQL endpoint.

• Since you have already checked the schema, please also review the Microsoft Fabric Known Issues page or your tenant’s Service Health dashboard for any ongoing SQL endpoint incidents.
Fabric known issues - Microsoft Fabric | Microsoft Learn

Thank you.

[v-sgandrathi]

Hi @debiagui,

Wanted to follow up and see if the provided response was useful. Glad to help

Thank you.

[Poojara_D12]

Hi @debiagui 

It sounds like you’re dealing with a case where newly inserted rows in your SQL database are not propagating to the SQL endpoint, even after several hours and after manually triggering a sync via the REST API. Normally, SQL endpoints rely on refresh or synchronization processes that should make recent changes available within a short timeframe, so a 5-hour delay indicates something abnormal. This could be due to delays in the underlying data refresh pipeline, indexing issues, or a known service incident that affects SQL endpoint updates. Since you’ve already tried a forced sync without success, the next steps would be to verify whether the ingestion pipeline is actually completing (check logs, triggers, or refresh history), confirm that there are no filtering or partitioning rules preventing the new rows from surfacing, and check the service health or known issues page for your SQL provider (for Microsoft Fabric or Azure Synapse, this would be the official documentation or service health dashboard). A good starting point is Microsoft’s Fabric SQL Endpoint documentation, as it includes details on expected sync behavior and troubleshooting steps.

Would you like me to also pull up the latest known issues or advisories for Fabric SQL endpoints so you can confirm if this is part of a wider service outage?