flume version : 1.7.0

canal version : 1.0.24

https://flume.apache.org/

Flume is a distributed, reliable, and available service for efficiently collecting, aggregating, and moving large amounts of log data. It has a simple and flexible architecture based on streaming data flows. It is robust and fault tolerant with tunable reliability mechanisms and many failover and recovery mechanisms. It uses a simple extensible data model that allows for online analytic application.

https://github.com/alibaba/canal

阿里巴巴mysql数据库binlog的增量订阅&消费组件

flume-canal-source 是对 flume 的 source 扩展。从 canal 获取数据到 flume channel。 进而可以实现binlog数据到 kafka / hdfs / hive / elasticsearch 等等。 **canal 和 flume 都有高可用的解决方案，这种方式同步 binlog 可用性非常高。**组合前人的优秀轮子，不重复造轮子。

部署 canal、flume 这里忽略。

• 配置 source 类型*

agent.sources = canalSource

agent.sources.canalSource.type = com.weiboyi.etl.flume.source.canal.CanalSource

• 配置连接 canal 的三种方式*

# 1. zookeeper servers
agent.sources.canalSource.zkServers = zookeeper-host:2181

# 2. canal server urls
agent.sources.canalSource.serverUrls = canal-server1:111111,canal-server2:111111

# 3. canal server urls
agent.sources.canalSource.serverUrl = canal-server1:111111

• 配置 canal destination*

agent.sources.canalSource.destination = example

• 配置用户名密码

agent.sources.canalSource.username = user
agent.sources.canalSource.password = passwd

• binlog batch size, default 1024

agent.sources.canalSource.batchSize = 1024

• 是否需要 MySQL 修改前的数据, default true

agent.sources.canalSource.oldDataRequired = true