Installing and Configuring Alibaba Canal for MySQL Binlog Capture

Installation

Download the Canal release package from the official GitHub repository and extract it to a fixed directory.

tar -zxvf canal.kafka-1.1.0.tar.gz

Configuration Files

canal.properties defines global settings such as instance list, configuration directory, auto‑scan options, and network parameters.

Parameter

Description

Default

canal.destinations

List of instance names deployed on the server

None

canal.conf.dir

Path to the conf/ directory

../conf

canal.auto.scan

Enable automatic scanning of instance configurations

true

canal.auto.scan.interval

Scanning interval in seconds

5

canal.instance.global.mode

Global configuration loading mode

spring

instance.properties contains instance‑specific settings such as MySQL connection details, charset, and filtering rules.

# Database address
canal.instance.master.address=192.168.56.104:3306

# Database user and password
canal.instance.dbUsername=canal
canal.instance.dbPassword=canal

# Connection charset
canal.instance.connectionCharset=UTF-8

# Default database
canal.instance.defaultDatabaseName=zgai_db

Additional global settings are defined in /usr/local/canal/conf/canal.properties (e.g., canal.id, canal.ip, canal.zkServers) and Kafka sink configuration in /usr/local/canal/conf/kafka.yml:

# Kafka servers
servers: 192.168.56.101:9092

# Batch size (KB)
canalBatchSize: 50

# Topic
topic: mytopic

Instance Management

After defining canal.destinations in canal.properties , create matching directories under conf/ and place an instance.properties file in each. If canal.auto.scan is enabled, Canal will automatically discover new, removed, or modified instance directories at the interval specified.

instance.xml Variants

Canal supports several instance.xml templates that determine how parser, sink, and store components persist state: spring/memory-instance.xml – all components in memory (fast, no HA). spring/file-instance.xml – file‑based persistence (single‑node, no HA). spring/default-instance.xml – Zookeeper + memory (supports HA). spring/group-instance.xml – logical grouping of multiple physical instances for sharded databases.

Key Parameter Highlights

canal.instance.mysql.slaveId

– unique ID for MySQL slave. canal.instance.master.address – MySQL master host. canal.instance.filter.regex – Perl‑style regex to select tables for capture. canal.instance.memory.buffer.size – Buffer size (must be power of two). canal.instance.detecting.enable – Enable heartbeat checks. canal.instance.fallbackIntervalInSeconds – Seconds to look back when MySQL master switches.

Understanding the relationship between parse position (tracked by CanalLogPositionManager) and consume position (tracked by CanalMetaManager) is essential for reliable data synchronization.

Choose the appropriate instance.xml based on deployment needs: memory for quick testing, file for simple production without HA, default (Zookeeper) for high‑availability clusters, and group for multi‑database aggregation.