Quick Start

This document describes how to quickly set up CCR (Cross Cluster Replication) in Apache Doris for cross-cluster data synchronization. It covers the full workflow from enabling binlog to submitting a sync task. After completing the four steps in order, you can establish a database-level or table-level synchronization link between a source cluster and a target cluster.

Applicable Scenarios

Scenario	Description
Database-level sync	Synchronize all tables in a database from the source cluster to the target cluster.
Table-level sync	Synchronize only one specific table from the source cluster to the target cluster.
Disaster recovery, read/write separation	Keep a real-time replica on the standby cluster through CCR. See Cross-Cluster Replication Overview.

Prerequisites

• A reachable source cluster and target cluster are deployed, and the Syncer has network connectivity to the upstream and downstream FE and BE.
• You have available accounts on both the source cluster and the target cluster (used by Syncer to start transactions and pull data). For permission requirements, see the Operations Manual.
• The Doris version meets the CCR requirements (2.0.15 / 2.1.6 or later is recommended).

Workflow Overview

1. Enable the binlog master switch on both the source cluster and the target cluster.
2. Download and deploy CCR Syncer.
3. Enable binlog on the database or table to be synchronized in the source cluster.
4. Submit a sync task to Syncer through the HTTP interface.

1. Enable Binlog on the Source and Target Clusters

Configure the following setting in fe.conf and be.conf on both the source cluster and the target cluster:

enable_feature_binlog=true

After modification, restart the corresponding FE and BE processes for the configuration to take effect.

2. Deploy Syncer

2.1 Download the Syncer Installation Package

Download the latest package from the following link:

https://download.selectdb.com/ccr-release/ccr-syncer-3.0.6-rc07-x64.tar.xz

2.2 Start and Stop Syncer

# Start
cd bin && sh start_syncer.sh --daemon

# Stop
sh stop_syncer.sh

After startup, Syncer listens on port 9190 by default to receive sync task management requests.

3. Enable Binlog on the Source Database / Table to Be Synchronized

Choose the enabling method based on the synchronization granularity:

• Whole-database sync: Run the script in Syncer's bin directory to enable binlog.enable for all tables in the specified database.

  ./enable_db_binlog.sh --host $host --port $port --user $user --password $password --db $db

• Single-table sync: Run ALTER TABLE on the target table in the source cluster to enable binlog.enable for that table.

  ALTER TABLE enable_binlog SET ("binlog.enable" = "true");

4. Submit a Sync Task to Syncer

Submit a sync task to Syncer through the HTTP interface:

curl -X POST -H "Content-Type: application/json" -d '{
    "name": "ccr_test",
    "src": {
      "host": "localhost",
      "port": "9030",
      "thrift_port": "9020",
      "user": "root",
      "password": "",
      "database": "your_db_name",
      "table": "your_table_name"
    },
    "dest": {
      "host": "localhost",
      "port": "9030",
      "thrift_port": "9020",
      "user": "root",
      "password": "",
      "database": "your_db_name",
      "table": "your_table_name"
    }
}' http://127.0.0.1:9190/create_ccr

Sync Task Parameters

Parameter	Description
name	Name of the CCR sync task. Must be unique; the same name can only be created once.
host, port	Host and MySQL (JDBC) port of the Master FE of the corresponding cluster.
thrift_port	Thrift port of the FE of the corresponding cluster.
user, password	The identity Syncer uses to start transactions, pull data, and so on.
database, table	For database-level sync, fill in your_db_name and leave your_table_name empty. For table-level sync, fill in both your_db_name and your_table_name.

FAQ

Q: What if enable_feature_binlog=true does not take effect after modification?

Confirm that both FE and BE on the source cluster and the target cluster have been restarted.

Q: What if some tables are not synchronized during whole-database sync?

Confirm that enable_db_binlog.sh has successfully enabled binlog.enable on every table in that database.

Q: What if creating a task with a duplicate name returns an error?

A name can only be used once. Use a different task name and try again.

Q: How can I learn more about configuration options and the feature matrix?

See Configuration and Feature Details.