本文适用语言:java

序章:定时任务实现方式

当下,java编码过程中,实现定时任务的方式主要以以下两种为主

网络上关于这两种框架的实践和配置相关的教程很多,这里不再赘述。
本文主要就二者的框架原理实现做一个入门引导,为了解深层实现细节做一定的铺垫。
本文源码版本

  • spring-context-3.2.18.RELEASE.jar
  • quartz-1.8.6.jar

一、Scheduled

1.1 使用方法

@EnableScheduling // @EnableScheduling 在配置类上使用,开启计划任务的支持
@Component(value="myClass")// 由spring管理
public class MyClass {

    @Scheduled(cron= "0 0 0 * * ?")//0 0 12 * * ? 每天12点触发0 0 0/1 * * ?  0 0 0 * * ?
    public void myTask() {
        // 业务逻辑
        ...
    }
}

1.2 源码分析

1.2.1 定时任务执行入口在哪?

org.springframework.scheduling.config.ContextLifecycleScheduledTaskRegistrar

public void onApplicationEvent(ContextRefreshedEvent event) {
	if (event.getApplicationContext() != this.applicationContext) {
		return;
	}
	// 定时任务执行入口方法绑定到容器生命周期上
	scheduleTasks();
}

1.2.2 调用链路

1. 所有已注册task
org.springframework.scheduling.config.ScheduledTaskRegistrar
protected void scheduleTasks() {
    ...
    if (this.triggerTasks != null) {
    	for (TriggerTask task : this.triggerTasks) {
            // 执行初始化完成的task和Trigger
    		this.scheduledFutures.add(this.taskScheduler.schedule(
    				task.getRunnable(), task.getTrigger()));
    	}
    }
    ...
}

2. 单个task
org.springframework.scheduling.TaskScheduler
ScheduledFuture schedule(Runnable task, Trigger trigger);

3. 线程池执行task
org.springframework.scheduling.concurrent.ThreadPoolTaskScheduler
public ScheduledFuture schedule(Runnable task, Trigger trigger) {
	ScheduledExecutorService executor = getScheduledExecutor();
	try {
		ErrorHandler errorHandler =
				(this.errorHandler != null ? this.errorHandler : TaskUtils.getDefaultErrorHandler(true));
		// 调用具体的实现方法.schedule()
		return new ReschedulingRunnable(task, trigger, executor, errorHandler).schedule();
	}
	catch (RejectedExecutionException ex) {
		throw new TaskRejectedException("Executor [" + executor + "] did not accept task: " + task, ex);
	}
}

4. 这块是具体的线程实现细节,已经与schedul无关
private <V> ScheduledFuture<V> schedule(final ScheduledFutureTask<V> task) {
    if (task == null) {
        throw new NullPointerException("task");
    } else {
        if (this.inEventLoop()) {
            this.delayedTaskQueue.add(task);
        } else {
            // 此处就是真正的线程执行方法
            this.execute(new Runnable() {
                public void run() {
                    SingleThreadEventExecutor.this.delayedTaskQueue.add(task);
                }
            });
        }

        return task;
    }
}

1.2.3 @Scheduled注解的生效原理

org.springframework.scheduling.annotation.ScheduledAnnotationBeanPostProcessor

// BeanPostProcessor生命周期方法,spring加载的时候会执行
public Object postProcessAfterInitialization(final Object bean, String beanName) {
		Class<?> targetClass = AopUtils.getTargetClass(bean);
	if (!this.nonAnnotatedClasses.containsKey(targetClass)) {
		final Set<Method> annotatedMethods = new LinkedHashSet<Method>(1);
		ReflectionUtils.doWithMethods(targetClass, new MethodCallback() {
			public void doWith(Method method) throws IllegalArgumentException, IllegalAccessException {
				Scheduled scheduled = AnnotationUtils.getAnnotation(method, Scheduled.class);
				if (scheduled != null) {
				    // @Scheduled的真正解析方法,具体解析细节和参数参看源码
				    // 解析后添加到ScheduledTaskRegistrar里
				    // 全部任务解析完成,执行ScheduledTaskRegistrar,具体实现参看[1.2.2 调用链路]章节
					processScheduled(scheduled, method, bean);
					annotatedMethods.add(method);
				}
			}
		});
		if (annotatedMethods.isEmpty()) {
			this.nonAnnotatedClasses.put(targetClass, Boolean.TRUE);
		}
	}
	return bean;
}

二、QUARTZ

2.1 使用方法

// 实例化一个调度器工厂,每个应用只有唯一一个工厂实例
SchedulerFactory schedFact = new org.quartz.impl.StdSchedulerFactory();
// 实例化一个调度器
Scheduler sched = schedFact.getScheduler();
// 启动,只有启动了调度器Quartz才会去执行任务
sched.start();

// 实例化一个任务
JobDetail job = newJob(HelloJob.class)
  .withIdentity("myJob", "group1")
  .build();

// 实例化一个任务触发器,立刻触发,每40s执行一次
Trigger trigger = newTrigger()
  .withIdentity("myTrigger", "group1")
  .startNow()
  .withSchedule(simpleSchedule()
      .withIntervalInSeconds(40)
      .repeatForever())
  .build();

// 调度任务
sched.scheduleJob(job, trigger);

2.2 源码分析

2.2.1 启动入口


1. web.xml配置
<context-param>
   <param-name>quartz:config-file</param-name>
   <param-value>/some/path/my_quartz.properties</param-value>
</context-param>
<context-param>
   <param-name>quartz:shutdown-on-unload</param-name>
   <param-value>true</param-value>
</context-param>
<context-param>
   <param-name>quartz:start-on-load</param-name>
   <param-value>true</param-value>
</context-param>

<listener>
   <listener-class>
       org.quartz.ee.servlet.QuartzInitializerListener
   </listener-class>
</listener>

2. org.quartz.ee.servlet.QuartzInitializerListener
// 执行ServletContextListener.contextInitialized的容器生命周期方法
public void contextInitialized(ServletContextEvent sce) {
    ...
    // 根据自定义的配置文件加载SchedulerFactory
    if (configFile != null) {
        factory = new StdSchedulerFactory(configFile);
    } else {
        factory = new StdSchedulerFactory();
    }
    
    // 加载scheduler
    scheduler = factory.getScheduler();
    
    // 启动scheduler
    scheduler.start();
    log.info("Scheduler has been started...");
    ...
}   

2.2.2 核心方法详解

1. StdSchedulerFactory.getScheduler()
public Scheduler getScheduler() throws SchedulerException {
    if (cfg == null) {
        // 根据不同的配置方式加载对应配置
        initialize();
    }
    ... 
    // 加载实例(加载Scheduler整个上下文环境)
    sched = instantiate();
    return sched;
}

2. StdSchedulerFactory.getScheduler().instantiate()
具体实现代码很多,以下做伪代码描述
private Scheduler instantiate() throws SchedulerException {

    // 校验初始化
    if (cfg == null) {
        initialize();
    }
    
    // 获取 Scheduler
    // 加载 ThreadPool
    // 加载 JobStore
    // 加载 DataSources
    // 加载 SchedulerPlugins
    // 加载 JobListeners
    // 加载 TriggerListeners
    // 加载 ThreadExecutor
    
    // 构造QuartzScheduler
    qs = new QuartzScheduler(rsrcs, schedCtxt, idleWaitTime, dbFailureRetry);
    Scheduler scheduler = instantiate(rsrcs, qs);
    qs.initialize();
    
    // 返回实例化好的scheduler
    return scheduler;
}