SpringBoot3项目实战：用ShardingJDBC 5.5.0搞定读写分离，附赠强制走主库的AOP小技巧

SpringBoot3实战ShardingJDBC 5.5.0读写分离与主库路由高阶技巧最近在重构一个订单管理系统时遇到了数据库性能瓶颈——高峰期查询请求导致主库负载飙升而多个从库却处于闲置状态。这让我意识到是时候引入读写分离架构了。经过技术选型最终选择了ShardingJDBC 5.5.0作为解决方案不仅因为它与SpringBoot3完美兼容更因其轻量级、无侵入的特性。本文将分享从基础配置到高级特性的完整实现路径特别是那些官方文档没有明确说明的实战细节。1. 环境准备与依赖配置在开始之前需要确认几个关键点SpringBoot3要求Java17环境而ShardingJDBC 5.5.0是首个全面支持SpringBoot3的版本。以下是经过生产验证的依赖组合dependency groupIdorg.apache.shardingsphere/groupId artifactIdshardingsphere-jdbc/artifactId version5.5.0/version exclusions exclusion groupIdorg.apache.shardingsphere/groupId artifactIdshardingsphere-test-util/artifactId /exclusion /exclusions /dependency关键排除项必须移除test-util依赖否则会与SpringBoot3的测试框架冲突。同时建议使用HikariCP连接池它在ShardingJDBC环境下表现最优spring: datasource: driver-class-name: org.apache.shardingsphere.driver.ShardingSphereDriver url: jdbc:shardingsphere:classpath:sharding.yaml注意allow-bean-definition-overriding必须设为true否则启动时会报Bean冲突2. 核心配置解剖sharding.yaml详解配置文件是ShardingJDBC的核心以下是经过优化的读写分离配置模板dataSources: master: driverClassName: com.mysql.cj.jdbc.Driver jdbcUrl: jdbc:mysql://master-host:3306/db username: user password: pass slave1: driverClassName: com.mysql.cj.jdbc.Driver jdbcUrl: jdbc:mysql://slave1-host:3306/db username: user password: pass rules: - !SINGLE tables: - *.* defaultDataSource: master - !READWRITE_SPLITTING dataSources: readwrite_ds: writeDataSourceName: master readDataSourceNames: - slave1 loadBalancerName: round_robin props: sql-show: true关键配置项解析!SINGLE规则必须配置且放在首位用于处理系统表查询负载均衡策略RANDOM随机选择默认ROUND_ROBIN轮询WEIGHT加权访问事务查询策略PRIMARY事务内强制走主库推荐FIXED固定走从库3. 高级特性注解式主库路由某些业务场景必须读取最新数据如支付状态校验这时需要强制走主库。我们设计了一套优雅的解决方案3.1 定义主库路由注解Retention(RetentionPolicy.RUNTIME) Target(ElementType.METHOD) public interface MasterRoute { String value() default ; }3.2 实现AOP切面Aspect Component RequiredArgsConstructor public class MasterRouteAspect { private final DataSource dataSource; Around(annotation(masterRoute)) public Object around(ProceedingJoinPoint joinPoint, MasterRoute masterRoute) throws Throwable { if (dataSource instanceof ShardingSphereDataSource) { try (HintManager hintManager HintManager.getInstance()) { hintManager.setWriteRouteOnly(); return joinPoint.proceed(); } } return joinPoint.proceed(); } }3.3 使用示例Service public class OrderService { MasterRoute public Order getLatestOrder(Long orderId) { // 该方法会强制走主库 return orderMapper.selectById(orderId); } }4. 性能优化与监控实施读写分离后需要建立完善的监控体系关键监控指标指标项监控方式健康阈值主从延迟SHOW SLAVE STATUS500ms从库负载均衡自定义负载统计偏差20%SQL执行耗时ShardingSphere的sql-show功能慢查询100ms启用SQL日志分析Configuration public class ShardingConfig { Bean public ShardingSphereDataSourceCustomizer dataSourceCustomizer() { return dataSource - { Properties props new Properties(); props.setProperty(sql-show, true); props.setProperty(sql-simple, false); dataSource.getContextManager().alterProperties(props); }; } }5. 常见问题解决方案问题1启动时报Table xxx doesnt exist解决方案检查!SINGLE规则是否正确定义且必须放在rules首位问题2事务内读不到刚写入的数据解决方案配置transactionalReadQueryStrategy: PRIMARY问题3从库负载不均衡优化方案调整负载策略为WEIGHT并配置权重loadBalancers: slave_weight: type: WEIGHT props: slave1: 2 slave2: 1在电商项目实践中这套方案将数据库查询性能提升了3倍主库负载下降60%。特别是在秒杀场景下通过MasterRoute精确控制关键路径的数据库访问既保证了数据一致性又充分利用了从库资源。