spring boot 多线程并发执行定时任务

程序员文章站 2022-05-01 16:30:18

...

Spring Boot默认情况下，所有定时任务会在一个线程中去执行，下面看测试代码，定义了三个测试Job：

@Component
@Slf4j
public class TestJob {

    @Scheduled(cron = "0/3 * * * * ? ")
    public void job1() {

        log.info("do job1 start");
        try {
            Thread.sleep(10000);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
        log.info("do job1 end");

    }

    @Scheduled(cron = "0/7 * * * * ? ")
    public void job2() {

        log.info("do job2");
    }

    @Scheduled(cron = "0/13 * * * * ? ")
    public void job3() {

        log.info("do job3");
    }
}

理论上，如果互不干扰的话，三个任务的触发时间点为：

job1触发：0,3,6,9,12……

job2触发：0,7,14,21,28,35,42,49,56

job3触发：0,13,26,39,52

实际执行结果：

2019-03-01 14:59:24  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 14:59:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 14:59:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 14:59:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 14:59:35  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 14:59:36  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 14:59:46  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 14:59:46  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 14:59:46  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 14:59:48  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 14:59:58  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 14:59:58  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 14:59:58  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:00:00  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:00:00  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:00:00  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:00:10  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:00:10  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:00:12  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:00:22  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:00:22  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:00:22  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:00:24  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:00:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:00:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:00:34  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:00:35  INFO 7868 --- [   scheduling-1] c.yclouds.service.demo.TestJob  : do job2

会看到他们的线程名都为schedule-1，观察line1-4，job1执行过程中，job2，job3触发时间到了，也没有执行。而是等job1结束后，立马执行一次（即时34秒的时候并不是它们的触发时间）。

上干货，并发执行多个定时任务，只需要新增一个线程池配置即可：

@Configuration
public class ScheduleConfig implements SchedulingConfigurer {

    @Override
    public void configureTasks(ScheduledTaskRegistrar scheduledTaskRegistrar) {

        int corePoolSize = 5;
        String nameFormat = "Schedule-Pool-%d";

        // 设置线程名
        ThreadFactory namedThreadFactory = new ThreadFactoryBuilder()
            .setNameFormat(nameFormat).build();
        // 初始化线程池
        ThreadPoolExecutor threadPool = new ScheduledThreadPoolExecutor(corePoolSize,
            namedThreadFactory);
        scheduledTaskRegistrar.setScheduler(threadPool);
    }
}

下面再来看执行结果：

2019-03-01 15:21:00  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:21:00  INFO 11204 --- [Schedule-Pool-4] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:00  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:21:07  INFO 11204 --- [Schedule-Pool-4] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:10  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:21:12  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:21:13  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:21:14  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:21  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:22  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:21:24  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:21:26  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:21:28  INFO 11204 --- [Schedule-Pool-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:34  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:21:35  INFO 11204 --- [Schedule-Pool-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:36  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:21:39  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:21:42  INFO 11204 --- [Schedule-Pool-4] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:46  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:21:48  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:21:49  INFO 11204 --- [Schedule-Pool-4] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:52  INFO 11204 --- [Schedule-Pool-2] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:21:56  INFO 11204 --- [Schedule-Pool-1] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:21:58  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 end
2019-03-01 15:22:00  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 start
2019-03-01 15:22:00  INFO 11204 --- [Schedule-Pool-1] c.yclouds.service.demo.TestJob  : do job3
2019-03-01 15:22:00  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:22:07  INFO 11204 --- [Schedule-Pool-3] c.yclouds.service.demo.TestJob  : do job2
2019-03-01 15:22:10  INFO 11204 --- [Schedule-Pool-0] c.yclouds.service.demo.TestJob  : do job1 end

三个任务都在各自的线程里执行，到点就触发，互不干扰。

上一篇： SQL SERVER存储过程的使用

下一篇： PHP中file_exists使用中遇到的问题小结，file_exists小结

spring boot 多线程并发执行定时任务

spring定时任务(scheduler)的串行、并行执行实现解析

Spring Boot @Scheduled定时任务代码实例解析

spring boot使用自定义的线程池执行Async任务

SpringBoot系列：Spring Boot集成定时任务Quartz

Spring Boot+Quartz实现一个实时管理的定时任务

Spring Boot实现简单的定时任务

Spring Boot @Async 异步任务执行

使用spring-task定时任务动态配置修改执行时间

Spring Boot与Kotlin定时任务的示例（Scheduling Tasks）

定时任务每次都执行两次的问题，慎用new ClassPathXmlApplicationContext()