mirroring-limitations.md

title	Limitations of Mirroring for SQL database
description	Details on the limitations of mirroring for SQL database in Fabric.
ms.reviewer	nzagorac, imotiwala, ajayj
ms.date	02/24/2026
ms.topic	concept-article

Limitations for Fabric SQL database mirroring

Current limitations in the Fabric SQL database mirroring are listed in this page. This page is subject to change.

For troubleshooting, see:

• Troubleshoot mirroring from Fabric SQL database
• Troubleshoot Fabric mirrored databases
• Troubleshoot Fabric mirrored databases from Azure SQL Database

For general limitations for SQL database in Microsoft Fabric, see Limitations in SQL database in Microsoft Fabric.

Database level limitations

• Mirroring for SQL database in Fabric cannot be disabled currently.
• Mirroring for SQL database can only mirror to the workspace in which the Fabric SQL database resides.
• The maximum number of tables that can be mirrored from one database is 1000. You can have more tables but any tables created beyond 1000 are skipped from mirroring.

Security features

• Row-level security is supported for Fabric SQL database, but permissions are currently not propagated to the replicated data in Fabric OneLake.
• Object-level permissions, for example granting permissions to certain columns, are currently not propagated to the replicated data in Fabric OneLake.
• Dynamic data masking settings are currently not propagated to the replicated data in Fabric OneLake.
• Microsoft Purview Information Protection/sensitivity labels are not cascaded and mirrored to Fabric OneLake. For more information, see Protect sensitive data in SQL database with Microsoft Purview protection policies.

Table level

• A table cannot be mirrored if the primary key includes an unsupported data type.
• Source tables that have any of the following features in use cannot be mirrored to Fabric OneLake.
  • When mirroring is active, Clustered columnstore indexes cannot be created on an existing table.
    • You can add a CCI to an existing table if you first stop mirroring, add the clustered columnstore index, then restart mirroring. However, the table will not be mirrored.
      • If Mirroring is running (it usually is), it can be stopped using the sqldatabase API and then re-started using the sqldatabase API. For instructions on how to stop and start mirroring with an API call, see Start and stop SQL database mirroring with the Fabric REST API.
    • Clustered columnstore indexes are supported when they are created at the same time the table is created using the inline index syntax, however, the new table cannot be mirrored.
  • Temporal history tables and ledger history tables
  • Always Encrypted
  • In-memory tables
  • Graph
  • External tables
• The following table-level data definition language (DDL) operations aren't allowed:
  • Switch/Split/Merge partition
  • Alter primary key
  • Altering tables to rebuild partitions with DATA COMPRESSION = ROW or PAGE is not allowed.
• When there is DDL change, a complete data snapshot is restarted for the changed table, and data is reseeded.
• Views are not mirrored to OneLake.
• Stored procedures are not mirrored to OneLake.
• ALTER INDEX ALL is not allowed on the table. Altering individual indexes referred to by name is allowed.
• For temporal tables, the data table is mirrored, but the history table is excluded from mirroring.
  • Upon adding system versioning (converting to Temporal) two existing tables, the existing history table is automatically excluded from mirroring (even if it was mirrored in the past).
  • Upon removing system versioning (splitting temporal data from its history table), the history table is treated as a standalone table and automatically added to mirroring.
• Full-text indexing is not supported and cannot be created in SQL database in Microsoft Fabric.
• The NotSupported replication status in the Replication monitor page contains status information specific to the table, often caused by an unsupported data type.
• Currently, a table cannot be mirrored if it has the json or vector data type.
  • Currently, you cannot ALTER a column to the vector or json data type in SQL database in Fabric.

Column level

• If the source table contains computed columns, these columns are skipped and cannot be mirrored.
• If the source table contains columns with one of these data types, these columns cannot be mirrored to Fabric OneLake. The following data types are unsupported for mirroring:
  • image
  • text/ntext
  • xml
  • rowversion/timestamp
  • sql_variant
  • User Defined Types (UDT)
  • geometry
  • geography
  • hierarchyid
• Delta lake supports only six digits of precision.
  • Columns of SQL type datetime2, with precision of 7 fractional second digits, do not have a corresponding data type with same precision in Delta files in Fabric OneLake. A precision loss happens if columns of this type are mirrored and seventh decimal second digit will be trimmed.
  • A table cannot be mirrored if the primary key is one of these data types: datetime2(7), datetimeoffset(7), time(7), where 7 is seven digits of precision.
  • The datetimeoffset(7) data type does not have a corresponding data type with same precision in Delta files in Fabric OneLake. A precision loss (loss of time zone and seventh time decimal) occurs if columns of this type are mirrored.
• Column names for a SQL table cannot contain spaces nor the following characters: , ; { } ( ) \n \t =.
• If one or more columns in the table is of type Large Binary Object (LOB) with a size greater than 1 MB, the column data is truncated to size of 1 MB in Fabric OneLake.

SQL analytics endpoint limitations

• The SQL analytics endpoint is the same as the Lakehouse SQL analytics endpoint. It is the same read-only experience. See SQL analytics endpoint limitations of the warehouse.

Related content

• Limitations in SQL database in Microsoft Fabric