详解基于Spring Boot与Spring Data JPA的多数据源配置
由于项目需要,最近研究了一下基于spring Boot与Spring Data JPA的多数据源配置问题。以下是传统的单数据源配置代码。这里使用的是Spring的Annotation在代码内部直接配置的方式,没有使用任何XML文件。
@Configuration
@EnableJpaRepositories(basePackages = "org.lyndon.repository")
@EnableTransactionManagement
@PropertySource("classpath:application.properties")
public class JpaConfig {
private static final String DATABASE_DRIVER = "db.driver";
private static final String DATABASE_URL = "db.url";
private static final String DATABASE_USER = "db.user";
private static final String DATABASE_PASSWORD = "db.password";
private static final String PACKAGES_TO_SCAN = "packages.to.scan";
private static final String HIBERNATE_DIALECT = "hibernate.dialect";
private static final String HIBERNATE_SHOW_SQL = "hibernate.show.sql";
@Resource
private Environment env;
@Bean
public DataSource dataSource() {
DruidDataSource source = new DruidDataSource();
source.setDriverClassName(env.getRequiredProperty(DATABASE_DRIVER));
source.setUrl(env.getRequiredProperty(DATABASE_URL));
source.setUsername(env.getRequiredProperty(DATABASE_USER));
source.setPassword(env.getRequiredProperty(DATABASE_PASSWORD));
return source;
}
@Bean
public LocalContainerEntityManagerFactoryBean entityManagerFactory() {
LocalContainerEntityManagerFactoryBean factory = new LocalContainerEntityManagerFactoryBean();
factory.setDataSource(dataSource());
factory.setPersistenceProviderClass(HibernatePersistenceProvider.class);
factory.setPackagesToScan(env.getRequiredProperty(PACKAGES_TO_SCAN));
factory.setJpaProperties(hibernateProperties());
factory.afterPropertiesSet();
return factory;
}
@Bean
public PlatformTransactionManager transactionManager() {
JpaTransactionManager manager = new JpaTransactionManager();
manager.setEntityManagerFactory(entityManagerFactory().getObject());
return manager;
}
@Bean
public HibernateExceptionTranslator hibernateExceptionTranslator() {
return new HibernateExceptionTranslator();
}
private Properties hibernateProperties() {
Properties properties = new Properties();
properties.put(HIBERNATE_DIALECT, env.getRequiredProperty(HIBERNATE_DIALECT));
properties.put(HIBERNATE_SHOW_SQL, env.getRequiredProperty(HIBERNATE_SHOW_SQL));
return properties;
}
}
但是这一配置是不能简单地扩展到多数据源配置的,因为Spring Boot默认会为开发人员做很多工作,而这些工作与多数据源的配置相冲突,因此需要修改原来的配置内容。网上有很多讲解多数据源配置的文章,但是这些文章大多使用的是XML配置的方式,而且没有使用Spring Boot等比较新的Spring框架技术(比如很多人使用实现AbstractRoutingDataSource这一Spring提供的抽象类的方式,还需要切面的支持,无疑是相当繁琐的),已经不适用于在最新的工程项目中使用了,因为最新的Spring框架可以为我们完成很多事情,我们只需要去适应新的方法即可。为此,我通过研究Spring的官方文档和不断调试,实现了新的多数据源配置的方法,在此贴出,仅供参考。
首先,Spring的JPA是直接支持多数据源配置的,因此我们可以在配置文件或者代码中直接配置多个数据源。由于多数据源配置可能会共享一些配置信息,因此使用继承体系实现这种配置是最合适的。在例子中,我的ORM框架使用的是hibernate,而与Hibernate相关的配置信息我都放在了配置基类BaseJpaConfg中,代码如下。
@Configuration
@EnableTransactionManagement
@PropertySource("classpath:application.properties")
public class BaseJpaConfig {
private static final String HIBERNATE_DIALECT = "hibernate.dialect";
private static final String HIBERNATE_SHOW_SQL = "hibernate.show.sql";
@Resource
protected Environment env;
@Bean
public HibernateExceptionTranslator hibernateExceptionTranslator() {
return new HibernateExceptionTranslator();
}
protected Properties hibernateProperties() {
Properties properties = new Properties();
properties.put(HIBERNATE_DIALECT, env.getRequiredProperty(HIBERNATE_DIALECT));
properties.put(HIBERNATE_SHOW_SQL, env.getRequiredProperty(HIBERNATE_SHOW_SQL));
return properties;
}
}
该类提供Hibernate相关信息。它有两个子类,分别是SpringJpaConfig以及MysqlJpaConfig,分别配置了两个数据源,连接到两个不同的数据库Schema,代码如下。
@Configuration
@EnableJpaRepositories(basePackages = {"org.lyndon.repository1"}, entityManagerFactoryRef = "springEntityManagerFactory",
transactionManagerRef = "springTransactionManager")
public class SpringJpaConfig extends BaseJpaConfig {
private static final String DATABASE_DRIVER = "db.driver";
private static final String DATABASE_URL = "db.url";
private static final String DATABASE_USER = "db.user";
private static final String DATABASE_PASSWORD = "db.password";
private static final String PACKAGES_TO_SCAN = "packages.to.scan";
@Bean
@ConfigurationProperties(prefix = "datasource.primary")
public DataSource dataSource() {
DruidDataSource source = new DruidDataSource();
source.setDriverClassName(env.getRequiredProperty(DATABASE_DRIVER));
source.setUrl(env.getRequiredProperty(DATABASE_URL));
source.setUsername(env.getRequiredProperty(DATABASE_USER));
source.setPassword(env.getRequiredProperty(DATABASE_PASSWORD));
return source;
}
@Bean(name = "springEntityManagerFactory")
public LocalContainerEntityManagerFactoryBean springEntityManagerFactory() {
LocalContainerEntityManagerFactoryBean factory = new LocalContainerEntityManagerFactoryBean();
factory.setDataSource(dataSource());
factory.setPersistenceProviderClass(HibernatePersistenceProvider.class);
factory.setPackagesToScan(env.getRequiredProperty(PACKAGES_TO_SCAN).split(","));
factory.setJpaProperties(hibernateProperties());
factory.afterPropertiesSet();
return factory;
}
@Bean(name = "springTransactionManager")
@Primary
public PlatformTransactionManager transactionManager() {
JpaTransactionManager manager = new JpaTransactionManager();
manager.setEntityManagerFactory(springEntityManagerFactory().getObject());
return manager;
}
}
@Configuration
@EnableJpaRepositories(basePackages = {"org.lyndon.repository2"}, entityManagerFactoryRef = "mysqlEntityManagerFactory",
transactionManagerRef = "mysqlTransactionManager")
public class MysqlJpaConfig extends BaseJpaConfig {
private static final String DATABASE_DRIVER = "db.driver";
private static final String DATABASE_URL = "db.url2";
private static final String DATABASE_USER = "db.user";
private static final String DATABASE_PASSWORD = "db.password";
private static final String PACKAGES_TO_SCAN = "packages.to.scan2";
@Bean
@ConfigurationProperties(prefix = "datasource.secondary")
public DataSource dataSource2() {
DruidDataSource source = new DruidDataSource();
source.setDriverClassName(env.getRequiredProperty(DATABASE_DRIVER));
source.setUrl(env.getRequiredProperty(DATABASE_URL));
source.setUsername(env.getRequiredProperty(DATABASE_USER));
source.setPassword(env.getRequiredProperty(DATABASE_PASSWORD));
return source;
}
@Bean(name = "mysqlEntityManagerFactory")
public LocalContainerEntityManagerFactoryBean mysqlEntityManagerFactory() {
LocalContainerEntityManagerFactoryBean factory = new LocalContainerEntityManagerFactoryBean();
factory.setDataSource(dataSource2());
factory.setPersistenceProviderClass(HibernatePersistenceProvider.class);
factory.setPackagesToScan(env.getRequiredProperty(PACKAGES_TO_SCAN).split(","));
factory.setJpaProperties(hibernateProperties());
factory.afterPropertiesSet();
return factory;
}
@Bean(name = "mysqlTransactionManager")
public PlatformTransactionManager transactionManager() {
JpaTransactionManager manager = new JpaTransactionManager();
manager.setEntityManagerFactory(mysqlEntityManagerFactory().getObject());
return manager;
}
}
两份配置代码使用了各自的数据源、实体管理对象以及事务管理对象。这里要注意的有两点,首先是实体管理对象。Spring Boot框架为方便开发者,默认会寻找名为“entityManagerFactory”的Bean作为实体管理的实现,但是我们这里使用了两个实体管理对象,名称也并不是默认的“entityManagerFactory”,因此必须在各自的EnableJpaRepositories的Annotation中指明使用的实体管理对象。第二点,就是事务管理对象。Spring Boot会提供一个默认的事务管理对象的实现,但是我们在这里使用了两个不同的事务管理对象,因此我们也需要在各自的EnableJpaRepositories的Annotation中指明使用的事务管理对象。除此以外,我们还必须指定其中一个事务管理对象为主要对象(使用Primary这一Annotation),让Spring能有主次地使用相应的事务管理对象。
以上是 详解基于Spring Boot与Spring Data JPA的多数据源配置 的全部内容, 来源链接: utcz.com/z/319760.html