@Async注解

异步调用介绍

  1. 异步调用
    异步调用就是在不阻塞主线程的情况下执行高耗时方法
  2. 常规异步
    通过开启新线程实现
  3. 在Springboot中启用异步方法
    需要4个注解
    @EnableAsync 开启异步,可以放在@Controller层上方,也可以放在Application类的上方
    @Component 注册异步组件
    @Async 标注异步方法
    @Autowired 注入异步组件
  4. 进行一次异步调用
    首先在一个Config类上标注开启异步
    然后创建一个异步的组件类,就跟Service,Controller 一样一样的,用Component标注,Service也行
    在类内创建一个异步方法,打上Async 标记。这个方法必须是实例方法。
    然后就跟注入Service一样一样的了。
  5. 异步事务
    在Async 方法上标注@Transactional是没用的。
    在Async 方法调用的Service上标注@Transactional 有效。
  6. 异步方法的内部调用
    异步方法不支持内部调用,也就是异步方法不能写在需要调用他的类的内部。
    比如Class A 有a,b,c。b有Async标注。此时a对b的异步调用是失效的。
  7. 为什么异步方法必须是实例方法
    因为static方法不能被Override。因为@Async 异步方法的实现原理是通过注入一个代理类到Bean中,这个代理继承这个Bean,需要覆写异步方法并执行。
    然后这个东西,会被Spring放到自己维护的一个队列中。等待线程池读取并执行。

线程池的使用

创建Service层的接口和实现

创建一个service层的接口AsyncService,如下:

public interface AsyncService {

    /**
     * 执行异步任务
     */
    void executeAsync();
}

对应的AsyncServiceImpl,实现如下:

@Service
public class AsyncServiceImpl implements AsyncService {

    private static final Logger logger = LoggerFactory.getLogger(AsyncServiceImpl.class);

    @Override
    public void executeAsync() {
        logger.info("start executeAsync");
        try{
            Thread.sleep(1000);
        }catch(Exception e){
            e.printStackTrace();
        }
        logger.info("end executeAsync");
    }
}

创建controller

创建一个controller为Hello,里面定义一个http接口,做的事情是调用Service层的服务,如下:

@RestController
public class Hello {

    private static final Logger logger = LoggerFactory.getLogger(Hello.class);

    @Autowired
    private AsyncService asyncService;

    @RequestMapping("/")
    public String submit(){
        logger.info("start submit");

        //调用service层的任务
        asyncService.executeAsync();

        logger.info("end submit");

        return "success";
    }
}

至此,我们已经做好了一个http请求的服务,里面做的事情其实是同步的,接下来我们就开始配置springboot的线程池服务,将service层做的事情都提交到线程池中去处理;

springboot的线程池配置

创建一个配置类ExecutorConfig,用来定义如何创建一个ThreadPoolTaskExecutor,要使用@Configuration@EnableAsync这两个注解,表示这是个配置类,并且是线程池的配置类,如下所示:

@Configuration
@EnableAsync
public class ExecutorConfig {

    private static final Logger logger = LoggerFactory.getLogger(ExecutorConfig.class);

    @Bean
    public Executor asyncServiceExecutor() {
        logger.info("start asyncServiceExecutor");
        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
        //配置核心线程数
        executor.setCorePoolSize(5);
        //配置最大线程数
        executor.setMaxPoolSize(5);
        //配置队列大小
        executor.setQueueCapacity(10);
        //配置线程池中的线程的名称前缀
        executor.setThreadNamePrefix("async-service-");

        // 设置拒绝策略:当pool已经达到max size的时候,如何处理新任务
        // CALLER_RUNS:不在新线程中执行任务,而是有调用者所在的线程来执行
        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
        //执行初始化
        executor.initialize();
        return executor;
    }
}

注意,上面的方法名称为asyncServiceExecutor,稍后马上用到;

将Service层的服务异步化

打开AsyncServiceImpl.java,在executeAsync方法上增加注解@Async(“asyncServiceExecutor”)asyncServiceExecutor是前面ExecutorConfig.java中的方法名,表明executeAsync方法进入的线程池是asyncServiceExecutor方法创建的,如下:

@Override
    @Async("asyncServiceExecutor")
    public void executeAsync() {
        logger.info("start executeAsync");
        try{
            Thread.sleep(1000);
        }catch(Exception e){
            e.printStackTrace();
        }
        logger.info("end executeAsync");
    }

验证效果

  1. 将这个springboot运行起来(pom.xml所在文件夹下执行mvn spring-boot:run);
  2. 在浏览器输入:http://localhost:8080
  3. 在浏览器用F5按钮快速多刷新几次;
  4. 在springboot的控制台看见日志如下:
2022-01-21 22:43:18.630  INFO 14824 --- [nio-8080-exec-8] c.b.t.controller.Hello                   : start submit
2022-01-21 22:43:18.630  INFO 14824 --- [nio-8080-exec-8] c.b.t.controller.Hello                   : end submit
2022-01-21 22:43:18.929  INFO 14824 --- [async-service-1] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:18.930  INFO 14824 --- [async-service-1] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync
2022-01-21 22:43:19.005  INFO 14824 --- [async-service-2] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:19.006  INFO 14824 --- [async-service-2] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync
2022-01-21 22:43:19.175  INFO 14824 --- [async-service-3] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:19.175  INFO 14824 --- [async-service-3] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync
2022-01-21 22:43:19.326  INFO 14824 --- [async-service-4] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:19.495  INFO 14824 --- [async-service-5] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:19.930  INFO 14824 --- [async-service-1] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:20.006  INFO 14824 --- [async-service-2] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 22:43:20.191  INFO 14824 --- [async-service-3] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync

如上日志所示,我们可以看到controller的执行线程是”nio-8080-exec-8”,这是tomcat的执行线程,而service层的日志显示线程名为“async-service-1”,显然已经在我们配置的线程池中执行了,并且每次请求中,controller的起始和结束日志都是连续打印的,表明每次请求都快速响应了,而耗时的操作都留给线程池中的线程去异步执行;

扩展ThreadPoolTaskExecutor

虽然我们已经用上了线程池,但是还不清楚线程池当时的情况,有多少线程在执行,多少在队列中等待呢?这里我创建了一个ThreadPoolTaskExecutor的子类,在每次提交线程的时候都会将当前线程池的运行状况打印出来,代码如下:

public class VisiableThreadPoolTaskExecutor extends ThreadPoolTaskExecutor {
    private static final Logger logger = LoggerFactory.getLogger(VisiableThreadPoolTaskExecutor.class);

    private void showThreadPoolInfo(String prefix){
        ThreadPoolExecutor threadPoolExecutor = getThreadPoolExecutor();

        if(null==threadPoolExecutor){
            return;
        }

        logger.info("{}, {},taskCount [{}], completedTaskCount [{}], activeCount [{}], queueSize [{}]",
                this.getThreadNamePrefix(),
                prefix,
                threadPoolExecutor.getTaskCount(),
                threadPoolExecutor.getCompletedTaskCount(),
                threadPoolExecutor.getActiveCount(),
                threadPoolExecutor.getQueue().size());
    }

    @Override
    public void execute(Runnable task) {
        showThreadPoolInfo("1\. do execute");
        super.execute(task);
    }

    @Override
    public void execute(Runnable task, long startTimeout) {
        showThreadPoolInfo("2\. do execute");
        super.execute(task, startTimeout);
    }

    @Override
    public Future<?> submit(Runnable task) {
        showThreadPoolInfo("1\. do submit");
        return super.submit(task);
    }

    @Override
    public <T> Future<T> submit(Callable<T> task) {
        showThreadPoolInfo("2\. do submit");
        return super.submit(task);
    }

    @Override
    public ListenableFuture<?> submitListenable(Runnable task) {
        showThreadPoolInfo("1\. do submitListenable");
        return super.submitListenable(task);
    }

    @Override
    public <T> ListenableFuture<T> submitListenable(Callable<T> task) {
        showThreadPoolInfo("2\. do submitListenable");
        return super.submitListenable(task);
    }
}

如上所示,showThreadPoolInfo方法中将任务总数、已完成数、活跃线程数,队列大小都打印出来了,然后Override了父类的execute、submit等方法,在里面调用showThreadPoolInfo方法,这样每次有任务被提交到线程池的时候,都会将当前线程池的基本情况打印到日志中;

修改ExecutorConfig.javaasyncServiceExecutor方法,将ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor()改为ThreadPoolTaskExecutor executor = new VisiableThreadPoolTaskExecutor(),如下所示:

		@Bean
    public Executor asyncServiceExecutor() {
        logger.info("start asyncServiceExecutor");
        //使用VisiableThreadPoolTaskExecutor
        ThreadPoolTaskExecutor executor = new VisiableThreadPoolTaskExecutor();
        //配置核心线程数
        executor.setCorePoolSize(5);
        //配置最大线程数
        executor.setMaxPoolSize(5);
        //配置队列大小
        executor.setQueueCapacity(99999);
        //配置线程池中的线程的名称前缀
        executor.setThreadNamePrefix("async-service-");

        // rejection-policy:当pool已经达到max size的时候,如何处理新任务
        // CALLER_RUNS:不在新线程中执行任务,而是有调用者所在的线程来执行
        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
        //执行初始化
        executor.initialize();
        return executor;
    }

再次启动该工程,再浏览器反复刷新http://localhost:8080,看到的日志如下:

2022-01-21 23:04:56.113  INFO 15580 --- [nio-8080-exec-1] c.b.t.e.VisiableThreadPoolTaskExecutor   : async-service-, 2. do submit,taskCount [99], completedTaskCount [85], activeCount [5], queueSize [9]
2022-01-21 23:04:56.113  INFO 15580 --- [nio-8080-exec-1] c.b.t.controller.Hello                   : end submit
2022-01-21 23:04:56.225  INFO 15580 --- [async-service-1] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 23:04:56.225  INFO 15580 --- [async-service-1] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync
2022-01-21 23:04:56.240  INFO 15580 --- [nio-8080-exec-2] c.b.t.controller.Hello                   : start submit
2022-01-21 23:04:56.240  INFO 15580 --- [nio-8080-exec-2] c.b.t.e.VisiableThreadPoolTaskExecutor   : async-service-, 2. do submit,taskCount [100], completedTaskCount [86], activeCount [5], queueSize [9]
2022-01-21 23:04:56.240  INFO 15580 --- [nio-8080-exec-2] c.b.t.controller.Hello                   : end submit
2022-01-21 23:04:56.298  INFO 15580 --- [async-service-2] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 23:04:56.298  INFO 15580 --- [async-service-2] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync
2022-01-21 23:04:56.372  INFO 15580 --- [nio-8080-exec-3] c.b.t.controller.Hello                   : start submit
2022-01-21 23:04:56.373  INFO 15580 --- [nio-8080-exec-3] c.b.t.e.VisiableThreadPoolTaskExecutor   : async-service-, 2. do submit,taskCount [101], completedTaskCount [87], activeCount [5], queueSize [9]
2022-01-21 23:04:56.373  INFO 15580 --- [nio-8080-exec-3] c.b.t.controller.Hello                   : end submit
2022-01-21 23:04:56.444  INFO 15580 --- [async-service-3] c.b.t.service.impl.AsyncServiceImpl      : end executeAsync
2022-01-21 23:04:56.445  INFO 15580 --- [async-service-3] c.b.t.service.impl.AsyncServiceImpl      : start executeAsync

注意这一行日志:2. do submit,taskCount [101], completedTaskCount [87], activeCount [5], queueSize [9]

这说明提交任务到线程池的时候,调用的是submit(Callable task)这个方法,当前已经提交了101个任务,完成了87个,当前有5个线程在处理任务,还剩9个任务在队列中等待,线程池的基本情况一路了然;

posted @ 2022-11-04 15:37  dongye95  阅读(908)  评论(0编辑  收藏  举报