JavaDataArchiver

Goal

In most of production scenario, data is growing infinitely as time goes by, which may cause DB performance drops dramatically. However massive cold data are barely used, this is the reason JavaDataArchiver came for.

JavaDataArchiver project is designed to archive data, it offers two modes to achieve this goal:

Plug&play: A standalone running server with simple configuration.
Built-in SDK: A more flexible way to built in your Java project

Feature List

Input:
- Mysql
Output:
- Mysql
- File
Support SSH tunnel to connect internal DB
Self-define datasource and archive configuration
Selective DB/Table/Column to archive
Customized cold data definition

TO-DO

Postgresql input/output support
Oracle input/output support
Amazon S3 output support
Aliyun OSS output support
GUI configuration

目标

构建一个通用的数据归档中间件

适配多种数据源
适配多种归档方式

何时归档

当面对以下需求时，可能需要归档：

数据集过大影响性能，通过归档拆分冷热数据
将数据从一个数据源迁移/复制到另一个数据源

适用范围

按照特定排序条件，不会随着时间变化而发生 排序变化 和 数据变化 的数据源（数据的不可变性）

典型数据

订单类数据（完成后的订单在一段时间后不再可变）
日志类数据（不可变）

生命周期(lifecycle)

初始化(initialize)
读取源数据(read)
拼装源数据(assemble)
转换(convert)
处理目标数据(process)
完成(finalize)

快速开始

配置数据源

在 conf/datasource 目录下添加 sample_src.properties

datasource.jdbc.driverClassName=com.mysql.jdbc.Driver
datasource.jdbc.url=jdbc:mysql://localhost:3306/db_sample_src?useUnicode=true&characterEncoding=UTF-8&autoReconnect=true&autoReconnectForPools=true&zeroDateTimeBehavior=convertToNull
datasource.jdbc.username=sample
datasource.jdbc.password=sample
datasource.jdbc.maxActive=10
datasource.jdbc.hibernate.dialect=org.hibernate.dialect.MySQL5InnoDBDialect

在 conf/datasource 目录下添加 sample_des.properties

datasource.jdbc.driverClassName=com.mysql.jdbc.Driver
datasource.jdbc.url=jdbc:mysql://localhost:3306/db_sample_des?useUnicode=true&characterEncoding=UTF-8&autoReconnect=true&autoReconnectForPools=true&zeroDateTimeBehavior=convertToNull
datasource.jdbc.username=sample
datasource.jdbc.password=sample
datasource.jdbc.maxActive=10
datasource.jdbc.hibernate.dialect=org.hibernate.dialect.MySQL5InnoDBDialect

配置归档任务

在 conf/archiver 目录下添加 sample.properties

archiver.scanner.interval=1000
archiver.src.datasource=sample_src
archiver.des.datasource=sample_des
archiver.src.table.name=sample_table_src
archiver.des.table.name=sample_table_des

fallstool/JavaDataArchiver