Spring-retry基本使用.docx

背景介绍在实际工作过程中，重试是一个经常使用的手段。比如MQ发送消息失败，会采取重试手段，比如工程中使用RPC请求外部服务,可能因为网络波动出现超时而采取重试手段......可以看见重试操作是非常常见的一种处理问题,系统设计的手段而在之前我们项目中处理重拾操作依赖MQ自身的重试机制,但是这种机制不是很灵活,如果某些功能没有使用MQ的话,那么就不是那么方便了,而本文介绍的Spring-Retry却能够以一种很优雅的方式解决这种问题,当然目前版本的Spring-retry还不是完美的,还是有待改进的.不过已经很不错了.基本使用例子1@Configuration@EnableRetrypublic class Application { @Bean public Service service() { return new Service(); }}@Serviceclass Service { @Retryable(RemoteAccessException.class) public void service() { // ... do something } @Recover public void recover(RemoteAccessException e) { // ... panic }}例子2@org.springframework.stereotype.Servicepublic class Service1 { @Retryable(value = {RemoteAccessException.class, RuntimeException.class}, maxAttempts = 2, backoff = @Backoff(value = 2000)) public void service() { System.out.println(do some things); // this exception will just trigger recover1, do not trigger recover3 throw new RemoteAccessException(remote access exception); // this exception will just trigger recover2// throw new RuntimeException(runtime exception);// System.out.println(do another things); } // 如果使用注解的话,这个recover貌似只能写在本类中,我测试了如果将recover方法写在 // recoverService中,好像找不到 @Recover public void recover1(RemoteAccessException e) { System.out.println(e.getMessage()); System.out.println(do recover operation1); } @Recover public void recover2(RuntimeException e) { System.out.println(e.getMessage()); System.out.println(do recover operation2); } @Recover public void recover3(RemoteAccessException e) { System.out.println(e.getMessage()); System.out.println(do recover operation3); }}例子3@Servicepublic class Service2 { public void test(){ final RetryTemplate retryTemplate = new RetryTemplate(); final SimpleRetryPolicy policy = new SimpleRetryPolicy(3, Collections.Class? extends Throwable, Boolean singletonMap(Exception.class, true)); FixedBackOffPolicy fixedBackOffPolicy = new FixedBackOffPolicy(); fixedBackOffPolicy.setBackOffPeriod(100); retryTemplate.setRetryPolicy(policy); retryTemplate.setBackOffPoli