Spring Boot 多租户架构实现：基于上下文自动传递的独立资源隔离方案

一、核心设计思想
通过线程上下文自动传递租户ID，结合动态数据源路由和中间件连接工厂，实现MySQL、Redis、RocketMQ的完全自动化资源隔离。关键设计如下：

二、关键实现方案

租户上下文管理（核心）

public class TenantContext {// 使用TransmittableThreadLocal支持异步传递private static final TransmittableThreadLocal<String> CONTEXT = new TransmittableThreadLocal<>();public static void set(String tenantId) {CONTEXT.set(tenantId);DynamicDataSourceContextHolder.push(tenantId); // 同步MySQL上下文}public static String get() {return CONTEXT.get();}public static void clear() {CONTEXT.remove();DynamicDataSourceContextHolder.poll(); // 清理MySQL上下文}
}

MySQL动态数据源配置

@Configuration
@MapperScan(basePackages = "com.example.mapper")
public class DataSourceConfig {@Bean@ConfigurationProperties(prefix = "spring.datasource.master")public DataSource masterDataSource() {return DataSourceBuilder.create().build();}@Bean@ConfigurationProperties(prefix = "spring.datasource.tenantA")public DataSource tenantADatasource() {return DataSourceBuilder.create().build();}@Bean@ConfigurationProperties(prefix = "spring.datasource.tenantB")public DataSource tenantBDatasource() {return DataSourceBuilder.create().build();}@Beanpublic DataSource routingDataSource(@Qualifier("masterDataSource") DataSource master,@Qualifier("tenantADatasource") DataSource tenantA,@Qualifier("tenantBDatasource") DataSource tenantB) {Map<Object, Object> targetDataSources = new HashMap<>();targetDataSources.put("tenantA", tenantA);targetDataSources.put("tenantB", tenantB);DynamicRoutingDataSource routingDataSource = new DynamicRoutingDataSource();routingDataSource.setDefaultTargetDataSource(master);routingDataSource.setTargetDataSources(targetDataSources);return routingDataSource;}
}public class DynamicRoutingDataSource extends AbstractRoutingDataSource {@Overrideprotected Object determineCurrentLookupKey() {return DynamicDataSourceContextHolder.getDataSourceKey();}
}

Redis连接工厂隔离

@Configuration
public class RedisConfig {@Beanpublic RedisConnectionFactory tenantARedisFactory() {RedisStandaloneConfiguration config = new RedisStandaloneConfiguration();config.setDatabase(1); // 租户A专用DBreturn new LettuceConnectionFactory(config);}@Beanpublic RedisConnectionFactory tenantBRedisFactory() {RedisStandaloneConfiguration config = new RedisStandaloneConfiguration();config.setDatabase(2); // 租户B专用DBreturn new LettuceConnectionFactory(config);}@Beanpublic RedisTemplate<String, Object> redisTemplate(@Qualifier("tenantARedisFactory") RedisConnectionFactory factory) {RedisTemplate<String, Object> template = new RedisTemplate<>();template.setConnectionFactory(factory);template.setKeySerializer(new TenantAwareStringRedisSerializer());return template;}
}public class TenantAwareStringRedisSerializer implements StringRedisSerializer {@Overridepublic String serialize(String string) {String tenantId = TenantContext.get();return (tenantId == null) ? string : tenantId + ":" + string;}
}

RocketMQ生产者/消费者自动绑定

@Configuration
public class RocketMQConfig {@Beanpublic DefaultMQProducer tenantProducer() throws MQClientException {DefaultMQProducer producer = new DefaultMQProducer("tenant_producer");producer.setNamesrvAddr("127.0.0.1:9876");producer.start();return producer;}@Beanpublic RocketMQListener<String> tenantListener() {return new TenantAwareListener();}
}public class TenantAwareListener implements RocketMQListener<String> {@Overridepublic void onMessage(String message) {String tenantId = TenantContext.get();// 自动路由到租户专属TopicString targetTopic = "topic_" + tenantId;// 处理消息...}
}

三、自动化上下文传播

请求拦截器

@Component
public class TenantInterceptor implements HandlerInterceptor {@Overridepublic boolean preHandle(HttpServletRequest request, HttpServletResponse response, Object handler) {String tenantId = request.getHeader("X-Tenant-ID");TenantContext.set(tenantId);return true;}@Overridepublic void afterCompletion(HttpServletRequest request,HttpServletResponse response,Object handler,Exception ex) {TenantContext.clear();}
}

异步任务支持

@EnableAsync
@Configuration
public class AsyncConfig implements AsyncConfigurer {@Overridepublic Executor getAsyncExecutor() {ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();executor.setTaskDecorator(new ContextCopyingDecorator());executor.initialize();return executor;}private static class ContextCopyingDecorator implements TaskDecorator {@Overridepublic Runnable decorate(Runnable runnable) {String tenantId = TenantContext.get();return () -> {try {TenantContext.set(tenantId);runnable.run();} finally {TenantContext.clear();}};}}
}

四、完整工作流程

请求入口
网关拦截请求 → 解析X-Tenant-ID → 设置到TenantContext

数据访问层

@Service
public class OrderService {@Autowiredprivate OrderMapper orderMapper;public Order getOrder(Long id) {// 自动路由到当前租户的数据源return orderMapper.selectById(id);}
}

Redis操作

@Service
public class CacheService {@Autowiredprivate RedisTemplate<String, Object> redisTemplate;public void setValue(String key, Object value) {// 自动添加租户前缀redisTemplate.opsForValue().set(key, value);}
}

消息发送

@Service
public class MessageService {@Autowiredprivate DefaultMQProducer producer;public void sendOrderCreated(Order order) {Message msg = new Message("order_created", order.toJson());producer.send(msg); // 自动绑定租户Topic}
}

五、技术优势对比

方案	上下文传递方式	中间件隔离级别	代码侵入性	线程安全
手动切换	显式代码调用	应用层	高	需要处理
本方案	线程上下文自动传播	物理隔离	低	原生支持
动态注解	AOP切面	逻辑隔离	中	需要配置

六、生产环境优化点

连接池监控

spring:datasource:tenantA:hikari:maximum-pool-size: 20leak-detection-threshold: 30000

RocketMQ消费者隔离

@RocketMQMessageListener(topic = "tenant_#{tenantId}_topic",consumerGroup = "tenant_#{tenantId}_group"
)
public class TenantConsumer implements RocketMQListener<String> {// 自动注入当前租户的Consumer
}

Redis集群支持

@Bean
public RedisClusterConfiguration tenantARedisCluster() {RedisClusterConfiguration config = new RedisClusterConfiguration();config.setClusterNodes(Arrays.asList("127.0.0.1:7001", "127.0.0.1:7002"));config.setPassword("tenantA@2024");return config;
}