Skip to main content
This feature is in preview and subject to change. To share feedback and/or issues, contact Support.
Logical data replication is only supported in CockroachDB self-hosted clusters. New in v25.1: The CREATE LOGICALLY REPLICATED statement starts logical data replication (LDR) on a table(s) that runs between a source and destination cluster in an active-active setup. CREATE LOGICALLY REPLICATED creates the new table on the destination cluster automatically and conducts a fast, offline initial scan. It accepts unidirectional or bidirectional on as an option to create either one of the setups automatically. Once the offline initial scan completes, the new table will come online and is ready to serve queries. In a bidirectional setup, the second LDR stream will also initialize after the offline initial scan completes.
If the table to be replicated contains user-defined types, you must use the CREATE LOGICAL REPLICATION STREAM statement instead. You can set up unidirectional or bidirectional LDR manually with CREATE LOGICAL REPLICATION STREAM.
This page is a reference for the CREATE LOGICALLY REPLICATED SQL statement, which includes information on its parameters and options. For a step-by-step guide to set up LDR, refer to the Set Up Logical Data Replication page.

​
Required privileges

CREATE LOGICALLY REPLICATED requires one of the following privileges: Use the GRANT SYSTEM statement:

​
Synopsis

<?xml version=“1.0” encoding=“UTF-8”?>

​
Parameters

​
Options

​
Examples

CREATE LOGICALLY REPLICATED will automatically create the specified source tables on the destination cluster. For unidirectional and bidirectional, run the statement to start LDR on the destination cluster that does not contain the tables.

​
Unidirectional

From the destination cluster of the LDR stream, run:
Include the following: For details on managing schema changes, conflicts, and jobs when LDR is running, refer to the Manage Logical Data Replication page.

​
Bidirectional

Both clusters will act as a source and destination in bidirectional LDR setups. To start the LDR jobs, you must run this statement from the destination cluster that does not contain the tables:
Include the following: For details on managing schema changes, conflicts, and jobs when LDR is running, refer to the Manage Logical Data Replication page.

​
Multiple tables

To include multiple tables in an LDR stream, add the fully qualified table names in a list format. Ensure that the table name in the source table list and destination table list are in the same order:
For details on managing schema changes, conflicts, and jobs when LDR is running, refer to the Manage Logical Data Replication page.

​
See more