Edit

Share via

Troubleshoot Fabric mirrored databases

  • Article

Scenarios, resolutions, and workarounds for Microsoft Fabric mirrored databases.

Resources

Review the troubleshooting section of frequently asked questions for each data source:

Review limitations documentation for each data source:

Changes to Fabric capacity

Scenario Description
Fabric capacity paused Mirroring is stopped and you cannot list or access the mirrored database item. Resume or reassign the capacity to your workspace.
Fabric capacity resumed Known limitation: When capacity is resumed from a paused state, the mirrored database status appears as Running, but mirroring doesn't start automatically. As a result, changes made in the source are not replicated to OneLake.
To resume mirroring, go to the mirrored database in the Fabric portal, select Configure replication, and click Apply change. Mirroring will continue from where it was paused.
Note if the capacity is paused for a long time, mirroring may not resume from its stopping point and will reseed data from the beginning. For example, this can occur if the transaction log for the database is full.
Fabric capacity scaling Mirroring continues. If you scale down the capacity, be aware that the OneLake storage for the mirrored data is free up to a limit based on the capacity size, thus scaling down the capacity may incur additional storage charge. Learn more from Cost of mirroring.
Fabric capacity throttled Wait until the overload state is over or update your capacity. Mirroring will continue once the capacity is restored. Learn more from Actions you can take to recover from overload situations.
Fabric trial capacity expired Mirroring is stopped. To retain your mirrored database, purchase Fabric capacity. Learn more from Fabric trial capacity expires.

Data doesn't appear to be replicating

If you observe a delay in the appearance of mirrored data, check the following:

  • Mirroring status: In the Fabric portal monitoring page of the mirrored database, check the status of mirrored database and specific tables, and the "Last completed" column which indicates the last time that Fabric refreshes the mirrored table from source. Empty means the table is not yet mirrored.

    If you enable the workspace monitoring, you can check the mirroring execution latency in addition, by querying the ReplicatorBatchLatency value from the mirrored database operation logs.

    For source types like Azure SQL Database, Azure SQL Managed Instance and Azure Database for PostgreSQL, follow the specific instruction to also check the source database configuration and status.

  • Data in OneLake: Mirroring continuously replicates your data into OneLake in Delta Lake table format. To validate if the data lands in OneLake properly, you can create a shortcut from the mirrored tables into a Lakehouse, then build notebooks with Spark queries to query the data. Learn more about Explore with notebooks.

  • Data in SQL analytics endpoint: You can query mirrored data through the SQL analytics endpoint of the mirrored database or a Lakehouse with a shortcut to the mirrored data. When you see a delay, validate the mirroring status and data in OneLake as mentioned above first. If the data shows up in OneLake but not in SQL analytics endpoint, it may be caused by a delay in metadata sync in SQL analytics endpoint.

    You can manually force a refresh of the automatic metadata scanning. On the page for the SQL analytics endpoint, select the Refresh button as shown in the following image. Wait for some time then query the data again to check.

    Screenshot from the Fabric portal of how to force a refresh for SQL analytics endpoint metadata scanning.

Stop replication

When you select Stop replication, OneLake files remain as is, but incremental replication stops. You can restart the replication at any time by selecting Start replication. You might want to do stop/start replication when resetting the state of replication, after source database changes, or as a troubleshooting tool.

Replicate source schema hierarchy

When you mirror data from various types of source databases, your source schema hierarchy is preserved in the mirrored database. It ensures that your data remains consistently organized across different services, allowing you to consume it using the same logic in SQL analytics endpoint, Spark Notebooks, semantic models, and other references to the data.

For mirrored databases created before this feature enabled, you see the source schema is flattened in the mirrored database, and schema name is encoded into the table name. If you want to reorganize tables with schemas, recreate your mirrored database.

If you use API to create/update mirrored database, set value for property defaultSchema which indicates whether to replicate the schema hierarchy from the source database. Refer to the definition samples in Microsoft Fabric mirroring public REST API.

Delta column mapping support

Mirroring supports replicating columns containing spaces or special characters in names (such as , ; { } ( ) \n \t =) from your source databases to the mirrored databases. Behind the scene, mirroring writes data into OneLake with Delta column mapping enabled.

For tables that are already under replication before this feature enabled, to include columns with special character in names, you need to update the mirrored database settings by removing and re-adding those tables, or stop and restart the mirrored database.

Take ownership of a mirrored database

Currently, mirrored database doesn't support ownership change. If a mirrored database stops working because the item owner has left the organization or it's no longer valid, you need to recreate the mirrored database.

Supported regions

The following are the Fabric regions that support database mirroring and open mirroring:

Americas:

  • Brazil South
  • Canada Central
  • Canada East
  • Central US
  • East US
  • East US2
  • North Central US
  • South Central US
  • West US
  • West US 2
  • West US 3

Asia Pacific:

  • East Asia
  • Southeast Asia
  • Australia East
  • Australia Southeast
  • Central India
  • South India
  • Japan East
  • Korea Central

Europe, the Middle East, and Africa:

  • North Europe
  • West Europe
  • France Central
  • Germany West Central
  • Italy North
  • Norway East
  • Poland Central
  • South Africa North
  • Sweden Central
  • Switzerland North
  • Switzerland West
  • UAE North
  • UK South
  • UK West

Troubleshoot

This section contains general Mirroring troubleshooting steps.

I can't connect to a source database

  1. Check your connection details are correct, server name, database name, username, and password.
  2. Check the server is not behind a firewall or private virtual network. Open the appropriate firewall ports.

No views are replicated

Currently, views are not supported. Only replicating regular tables are supported.

No tables are being replicated

  1. Check the monitoring status to check the status of the tables. For more information, see Monitor Fabric mirrored database replication.
  2. Select the Configure replication button. Check to see if the tables are present in the list of tables, or if any Alerts on each table detail are present.

Columns are missing from the destination table

  1. Select the Configure replication button.
  2. Select the Alert icon next to the table detail if any columns are not being replicated.

Some of the data in my column appears to be truncated

The Fabric warehouse does not support VARCHAR(max) it only currently supports VARCHAR(8000).

I can't change the source database

Changing the source database is not supported. Create a new mirrored database.

Limits error messages

These common error messages have explanations and mitigations:

Error message Reason Mitigation
"The tables count may exceed the limit, there could be some tables missing." There's a maximum of 500 tables. In the source database, drop or filter tables. If the new table is the 500th table, no mitigation required.
"The replication is being throttled and expected to continue at YYYY-MM-DDTHH:MM:ss." There's a maximum of 1 TB of change data captured per Mirrored database per day. Wait for throttling to end.

Related content