作者 | 付政委
博客 | https://bugstack.cn
微信公众号：bugstack虫洞栈 | 博客：https://bugstack.cn 沉淀、分享、成长，专注于原创专题案例，以最易学习编程的方式分享知识，让自己和他人都能有所收获。目前已完成的专题有；Netty4.x实战专题案例、用Java实现JVM、基于JavaAgent的全链路监控、手写RPC框架、架构设计专题案例[Ing]等。 你用剑?、我用刀?，好的代码都很烧?，望你不吝出招?！
一、前言介绍
在日常开发中经常会用到定时任务，用来；库表扫描发送MQ、T n账单结算、缓存数据更新、秒杀活动状态变更，等等。因为有了Spring的Schedule极大的方便了我们对这类场景的使用。那么，除了应用你还了解它多少呢；
- 默认初始化多少个任务线程
- JobStore有几种实现，你平时用的都是哪个
- 一个定时任务的执行流程简述下
蒙圈了吧，是不感觉平时只是使用了，根本没关注过这些。有种冲动赶紧搜索答案吧！但只是知道答案是没有多少意义的，扛不住问不说，也不了解原理。所以，如果你想真的提升自己技能，还是要从根本搞定。
二、案例工程
为了更好的做源码分析，我们将平时用的定时任务服务单独抽离出来。工程下载，关注公众号：bugstack虫洞栈，回复：源码分析
代码语言：javascript复制 1itstack-demo-code-schedule
2└── src
3 ├── main
4 │ ├── java
5 │ │ └── org.itstack.demo
6 │ │ ├── DemoTask.java
7 │ │ └── JobImpl.java
8 │ └── resources
9 │ ├── props
10 │ │ └── config.properties
11 │ ├── spring
12 │ │ └── spring-config-schedule-task.xml
13 │ ├── logback.xml
14 │ └── spring-config.xml
15 └── test
16 └── java
17 └── org.itstack.demo.test
18 ├── ApiTest.java
19 ├── MyQuartz.java
20 └── MyTask.java三、环境配置
- JDK 1.8
- IDEA 2019.3.1
- Spring 4.3.24.RELEASE
- quartz 2.3.2 ｛不同版本略有代码差异｝
四、源码分析
代码语言：javascript复制1<dependency>
2 <groupId>org.quartz-scheduler</groupId>
3 <artifactId>quartz</artifactId>
4 <version>2.3.2</version>
5</dependency>依赖于Spring版本升级quartz选择2.3.2，同时如果你如本文案例中所示使用xml配置任务。那么会有如下更改；
代码语言：javascript复制Spring 3.x/org.springframework.scheduling.quart.CronTriggerBean
1 <bean id="taskTrigger" class="org.springframework.scheduling.quartz.CronTriggerBean">
2 <property name="jobDetail" ref="taskHandler"/>
3 <property name="cronExpression" value="0/5 * * * * ?"/>
4 </bean>代码语言：javascript复制Spring 4.x/org.springframework.scheduling.quartz.CronTriggerFactoryBean
1 <bean id="taskTrigger" class="org.springframework.scheduling.quartz.CronTriggerFactoryBean">
2 <property name="jobDetail" ref="taskHandler"/>
3 <property name="cronExpression" value="0/5 * * * * ?"/>
4 </bean>在正式分析前，可以看下quartz的默认配置，很多初始化动作都要从这里取得参数，同样你可以配置自己的配置文件。例如，当你的任务很多时，默认初始化的10个线程组不满足你的业务需求，就可以按需调整。
代码语言：javascript复制quart.properties
1# Default Properties file for use by StdSchedulerFactory
2# to create a Quartz Scheduler Instance, if a different
3# properties file is not explicitly specified.
4#
5
6org.quartz.scheduler.instanceName: DefaultQuartzScheduler
7org.quartz.scheduler.rmi.export: false
8org.quartz.scheduler.rmi.proxy: false
9org.quartz.scheduler.wrapJobExecutionInUserTransaction: false
10
11org.quartz.threadPool.class: org.quartz.simpl.SimpleThreadPool
12org.quartz.threadPool.threadCount: 10
13org.quartz.threadPool.threadPriority: 5
14org.quartz.threadPool.threadsInheritContextClassLoaderOfInitializingThread: true
15
16org.quartz.jobStore.misfireThreshold: 60000
17
18org.quartz.jobStore.class: org.quartz.simpl.RAMJobStore1. 从一个简单案例开始
平时我们使用Schedule基本都是注解或者xml配置文件，但是为了可以更简单的分析代码，我们从一个简单的Demo入手，放到main函数中。
代码语言：javascript复制DemoTask.java & 定义一个等待被执行的任务
1public class DemoTask {
2
3 private Logger logger = LoggerFactory.getLogger(DemoTask.class);
4
5 public void execute() throws Exception{
6 logger.info("定时处理用户信息任务：0/5 * * * * ?");
7 }
8
9}代码语言：javascript复制MyTask.java & 测试类，将配置在xml中的代码抽离出来
1public class MyTask {
2
3 public static void main(String[] args) throws Exception {
4
5 DemoTask demoTask = new DemoTask();
6
7 // 定义了；执行的内容
8 MethodInvokingJobDetailFactoryBean methodInvokingJobDetailFactoryBean = new MethodInvokingJobDetailFactoryBean();
9 methodInvokingJobDetailFactoryBean.setTargetObject(demoTask);
10 methodInvokingJobDetailFactoryBean.setTargetMethod("execute");
11 methodInvokingJobDetailFactoryBean.setConcurrent(true);
12 methodInvokingJobDetailFactoryBean.setName("demoTask");
13 methodInvokingJobDetailFactoryBean.afterPropertiesSet();
14
15 // 定义了；执行的计划
16 CronTriggerFactoryBean cronTriggerFactoryBean = new CronTriggerFactoryBean();
17 cronTriggerFactoryBean.setJobDetail(methodInvokingJobDetailFactoryBean.getObject());
18 cronTriggerFactoryBean.setCronExpression("0/5 * * * * ?");
19 cronTriggerFactoryBean.setName("demoTask");
20 cronTriggerFactoryBean.afterPropertiesSet();
21
22 // 实现了；执行的功能
23 SchedulerFactoryBean schedulerFactoryBean = new SchedulerFactoryBean();
24 schedulerFactoryBean.setTriggers(cronTriggerFactoryBean.getObject());
25 schedulerFactoryBean.setAutoStartup(true);
26 schedulerFactoryBean.afterPropertiesSet();
27
28 schedulerFactoryBean.start();
29
30 // 暂停住
31 System.in.read();
32
33 }
34
35}如果一切顺利，那么会有如下结果：
代码语言：javascript复制 12020-01-04 10:47:16.369 [main] INFO org.quartz.impl.StdSchedulerFactory[1220] - Using default implementation for ThreadExecutor
22020-01-04 10:47:16.421 [main] INFO org.quartz.core.SchedulerSignalerImpl[61] - Initialized Scheduler Signaller of type: class org.quartz.core.SchedulerSignalerImpl
32020-01-04 10:47:16.422 [main] INFO org.quartz.core.QuartzScheduler[229] - Quartz Scheduler v.2.3.2 created.
42020-01-04 10:47:16.423 [main] INFO org.quartz.simpl.RAMJobStore[155] - RAMJobStore initialized.
52020-01-04 10:47:16.424 [main] INFO org.quartz.core.QuartzScheduler[294] - Scheduler meta-data: Quartz Scheduler (v2.3.2) 'QuartzScheduler' with instanceId 'NON_CLUSTERED'
6 Scheduler class: 'org.quartz.core.QuartzScheduler' - running locally.
7 NOT STARTED.
8 Currently in standby mode.
9 Number of jobs executed: 0
10 Using thread pool 'org.quartz.simpl.SimpleThreadPool' - with 10 threads.
11 Using job-store 'org.quartz.simpl.RAMJobStore' - which does not support persistence. and is not clustered.
12
132020-01-04 10:47:16.424 [main] INFO org.quartz.impl.StdSchedulerFactory[1374] - Quartz scheduler 'QuartzScheduler' initialized from an externally provided properties instance.
142020-01-04 10:47:16.424 [main] INFO org.quartz.impl.StdSchedulerFactory[1378] - Quartz scheduler version: 2.3.2
152020-01-04 10:47:16.426 [main] INFO org.quartz.core.QuartzScheduler[2293] - JobFactory set to: org.springframework.scheduling.quartz.AdaptableJobFactory@3e9b1010
162020-01-04 10:47:16.651 [main] INFO org.quartz.core.QuartzScheduler[547] - Scheduler QuartzScheduler_$_NON_CLUSTERED started.
17一月 04, 2020 10:47:16 上午 org.springframework.scheduling.quartz.SchedulerFactoryBean startScheduler
18信息: Starting Quartz Scheduler now
192020-01-04 10:47:20.321 [QuartzScheduler_Worker-1] INFO org.itstack.demo.DemoTask[11] - 定时处理用户信息任务：0/5 * * * * ?
202020-01-04 10:47:25.001 [QuartzScheduler_Worker-2] INFO org.itstack.demo.DemoTask[11] - 定时处理用户信息任务：0/5 * * * * ?
212020-01-04 10:47:30.000 [QuartzScheduler_Worker-3] INFO org.itstack.demo.DemoTask[11] - 定时处理用户信息任务：0/5 * * * * ?
222020-01-04 10:47:35.001 [QuartzScheduler_Worker-4] INFO org.itstack.demo.DemoTask[11] - 定时处理用户信息任务：0/5 * * * * ?
232020-01-04 10:47:40.000 [QuartzScheduler_Worker-5] INFO org.itstack.demo.DemoTask[11] - 定时处理用户信息任务：0/5 * * * * ?
24
25Process finished with exit code -1
262. 定义执行内容(MethodInvokingJobDetailFactoryBean)
代码语言：javascript复制1// 定义了；执行的内容
2MethodInvokingJobDetailFactoryBean methodInvokingJobDetailFactoryBean = new MethodInvokingJobDetailFactoryBean();
3methodInvokingJobDetailFactoryBean.setTargetObject(demoTask);
4methodInvokingJobDetailFactoryBean.setTargetMethod("execute");
5methodInvokingJobDetailFactoryBean.setConcurrent(true);
6methodInvokingJobDetailFactoryBean.setName("demoTask");
7methodInvokingJobDetailFactoryBean.afterPropertiesSet();这块内容主要将我们的任务体(即待执行任务DemoTask)交给MethodInvokingJobDetailFactoryBean管理，首先设置必要信息；
- targetObject：目标对象bean，也就是demoTask
- targetMethod：目标方法name，也就是execute
- concurrent：是否并行执行，非并行执行任务，如果上一个任务没有执行完，下一刻不会执行
- name：xml配置非必传，源码中可以获取beanName
最后我们通过手动调用 afterPropertiesSet() 来模拟初始化。如果我们的类是交给 Spring 管理的，那么在实现了 InitializingBean 接口的类，在类配置信息加载后会自动执行 afterPropertiesSet() 。一般实现了 InitializingBean 接口的类，同时也会去实现 FactoryBean 接口，因为这个接口实现后就可以通过 T getObject() 获取自己自定义初始化的类。这也常常用在一些框架开发中。
代码语言：javascript复制MethodInvokingJobDetailFactoryBean.afterPropertiesSet()
1public void afterPropertiesSet() throws ClassNotFoundException, NoSuchMethodException {
2 prepare();
3 // Use specific name if given, else fall back to bean name.
4 String name = (this.name != null ? this.name : this.beanName);
5 // Consider the concurrent flag to choose between stateful and stateless job.
6 Class<?> jobClass = (this.concurrent ? MethodInvokingJob.class : StatefulMethodInvokingJob.class);
7 // Build JobDetail instance.
8 JobDetailImpl jdi = new JobDetailImpl();
9 jdi.setName(name);
10 jdi.setGroup(this.group);
11 jdi.setJobClass((Class) jobClass);
12 jdi.setDurability(true);
13 jdi.getJobDataMap().put("methodInvoker", this);
14 this.jobDetail = jdi;
15
16 postProcessJobDetail(this.jobDetail);
17}- 源码168行： 根据是否并行执行选择任务类，这两个类都是MethodInvokingJobDetailFactoryBean的内部类，非并行执行的StatefulMethodInvokingJob只是继承MethodInvokingJob添加了标记注解。
- 源码171行： 创建JobDetailImpl，添加任务明细信息，注意这类的jdi.setJobClass((Class) jobClass)实际就是MethodInvokingJob。MethodInvokingJob也是我们最终要反射调用执行的内容。
- 源码177行： 初始化任务后赋值给this.jobDetail = jdi，也就是最终的类对象 MethodInvokingJobDetailFactoryBean.getObject() 1@Override 2public JobDetail getObject() { 3 return this.jobDetail; 4}
- 源码：220行： 获取对象时返回 this.jobDetail，这也就解释了为什么 MethodInvokingJobDetailFactoryBean 初始化后直接赋值给了一个 JobDetail ；
微信公众号：bugstack虫洞栈 & Schedule.xml
3. 定义执行计划(CronTriggerFactoryBeann)
代码语言：javascript复制1// 定义了；执行的计划
2CronTriggerFactoryBean cronTriggerFactoryBean = new CronTriggerFactoryBean();
3cronTriggerFactoryBean.setJobDetail(methodInvokingJobDetailFactoryBean.getObject());
4cronTriggerFactoryBean.setCronExpression("0/5 * * * * ?");
5cronTriggerFactoryBean.setName("demoTask");
6cronTriggerFactoryBean.afterPropertiesSet();这一块主要定义任务的执行计划，并将任务执行内容交给 CronTriggerFactoryBean 管理，同时设置必要信息；
- jobDetail：设置任务体，xml 中可以直接将对象赋值，硬编码中设置执行的 JobDetail 对象信息。也就是我们上面设置的 JobDetailImpl ，通过 getObject() 获取出来。
- cronExpression：计划表达式；秒、分、时、日、月、周、年
代码语言：javascript复制CronTriggerFactoryBean.afterPropertiesSet()
1@Override
2public void afterPropertiesSet() throws ParseException {
3
4 // ... 校验属性信息
5
6 CronTriggerImpl cti = new CronTriggerImpl();
7 cti.setName(this.name);
8 cti.setGroup(this.group);
9 if (this.jobDetail != null) {
10 cti.setJobKey(this.jobDetail.getKey());
11 }
12 cti.setJobDataMap(this.jobDataMap);
13 cti.setStartTime(this.startTime);
14 cti.setCronExpression(this.cronExpression);
15 cti.setTimeZone(this.timeZone);
16 cti.setCalendarName(this.calendarName);
17 cti.setPriority(this.priority);
18 cti.setMisfireInstruction(this.misfireInstruction);
19 cti.setDescription(this.description);
20 this.cronTrigger = cti;
21}- 源码237行： 创建触发器 CronTriggerImpl 并设置相关属性信息
- 源码245行： 生成执行计划类 cti.setCronExpression(this.cronExpression);
1public void setCronExpression(String cronExpression) throws ParseException { 2 TimeZone origTz = getTimeZone(); 3 this.cronEx = new CronExpression(cronExpression); 4 this.cronEx.setTimeZone(origTz); 5}
CronExpression.java & 解析Cron表达式
1protected void buildExpression(String expression) throws ParseException { 2 expressionParsed = true; 3 try { // ... 初始化 TreeSet xxx = new TreeSet<Integer>(); 4 5 int exprOn = SECOND; 6 StringTokenizer exprsTok = new StringTokenizer(expression, " t", 7 false); 8 9 while (exprsTok.hasMoreTokens() && exprOn <= YEAR) { 10 String expr = exprsTok.nextToken().trim(); 11 12 // ... 校验DAY_OF_MONTH和DAY_OF_WEEK字段的特殊字符 13 14 StringTokenizer vTok = new StringTokenizer(expr, ","); 15 while (vTok.hasMoreTokens()) { 16 String v = vTok.nextToken(); 17 storeExpressionVals(0, v, exprOn); 18 } 19 exprOn ; 20 } 21 22 // ... 校验DAY_OF_MONTH和DAY_OF_WEEK字段的特殊字符 23 24} catch (ParseException pe) { 25 throw pe; 26} catch (Exception e) { 27 throw new ParseException("Illegal cron expression format (" 28 e.toString() ")", 0); 29} 30}
- Cron表达式有7个字段，CronExpression 把7个字段解析为7个 TreeSet 对象。
- 填充TreeSet对象值的时候，表达式都会转换为起始值、结束值和增量的计算模式，然后计算出匹配的值放进TreeSet对象
代码语言：javascript复制CronTriggerFactoryBean.getObject()
1@Override
2public CronTrigger getObject() {
3 return this.cronTrigger;
4}- 源码257行： 获取对象时返回 this.cronTrigger ，也就是 CronTriggerImpl 对象
4. 调度执行计划(SchedulerFactoryBean)
代码语言：javascript复制1// 调度了；执行的计划(scheduler)
2SchedulerFactoryBean schedulerFactoryBean = new SchedulerFactoryBean();
3schedulerFactoryBean.setTriggers(cronTriggerFactoryBean.getObject());
4schedulerFactoryBean.setAutoStartup(true);
5schedulerFactoryBean.afterPropertiesSet();
6
7schedulerFactoryBean.start();这一部分如名字一样调度工厂，相当于一个指挥官，可以从全局做调度，比如监听哪些trigger已经ready、分配线程等等，同样也需要设置必要的属性信息；
- triggers：按需可以设置多个触发器，本文设置了一个 cronTriggerFactoryBean.getObject() 也就是 CronTriggerImpl 对象
- autoStartup：默认是否自动启动任务，默认值为true
这个过程较长包括：调度工厂、线程池、注册任务等等，整体核心加载流程如下；
微信公众号：bugstack虫洞栈 & 调度工程初始化流程
- 整个加载过程较长，抽取部分核心代码块进行分析，其中包括的类；
- StdScheduler
- StdSchedulerFactory
- SimpleThreadPool
- QuartzScheduler
- QuartzSchedulerThread
- RAMJobStore
- CronTriggerImpl
- CronExpression
代码语言：javascript复制SchedulerFactoryBean.afterPropertiesSet()
1public void afterPropertiesSet() throws Exception {
2 if (this.dataSource == null && this.nonTransactionalDataSource != null) {
3 this.dataSource = this.nonTransactionalDataSource;
4 }
5 if (this.applicationContext != null && this.resourceLoader == null) {
6 this.resourceLoader = this.applicationContext;
7 }
8 // Initialize the Scheduler instance...
9 this.scheduler = prepareScheduler(prepareSchedulerFactory());
10 try {
11 registerListeners();
12 registerJobsAndTriggers();
13 }
14 catch (Exception ex) {
15 try {
16 this.scheduler.shutdown(true);
17 }
18 catch (Exception ex2) {
19 logger.debug("Scheduler shutdown exception after registration failure", ex2);
20 }
21 throw ex;
22 }
23}- 源码474行： 为调度器做准备工作 prepareScheduler(prepareSchedulerFactory()) ，依次执行如下；
11）初始化threadPool(线程池)：开发者可以通过org.quartz.threadPool.class配置指定使用哪个线程池类，比如SimpleThreadPool。 22）初始化jobStore(任务存储方式)：开发者可以通过org.quartz.jobStore.class配置指定使用哪个任务存储类，比如RAMJobStore。 33）初始化dataSource(数据源)：开发者可以通过org.quartz.dataSource配置指定数据源详情，比如哪个数据库、账号、密码等。 44）初始化其他配置：包括SchedulerPlugins、JobListeners、TriggerListeners等； 55）初始化threadExecutor(线程执行器)：默认为DefaultThreadExecutor； 66）创建工作线程：根据配置创建N个工作thread，执行start()启动thread，并将N个thread顺序add进threadPool实例的空闲线程列表availWorkers中； 77）创建调度器线程：创建QuartzSchedulerThread实例，并通过threadExecutor.execute(实例)启动调度器线程； 88）创建调度器：创建StdScheduler实例，将上面所有配置和引用组合进实例中，并将实例存入调度器池中
- SchedulerFactoryBean.prepareScheduler(SchedulerFactory schedulerFactory)
- SchedulerFactoryBean.createScheduler(schedulerFactory, this.schedulerName);
- SchedulerFactoryBean.createScheduler(SchedulerFactory schedulerFactory, String schedulerName)
- Scheduler newScheduler = schedulerFactory.getScheduler();
- StdSchedulerFactory.getScheduler();
- sched = instantiate(); 包括一系列核心操作；
- 源码477行： 调用父类 SchedulerAccessor.registerJobsAndTriggers() 注册任务和触发器 1for (Trigger trigger : this.triggers) { 2 addTriggerToScheduler(trigger); 3}
代码语言：javascript复制SchedulerAccessor.addTriggerToScheduler() & SchedulerAccessor 是SchedulerFactoryBean的父类
1private boolean addTriggerToScheduler(Trigger trigger) throws SchedulerException {
2 boolean triggerExists = (getScheduler().getTrigger(trigger.getKey()) != null);
3 if (triggerExists && !this.overwriteExistingJobs) {
4 return false;
5 }
6 // Check if the Trigger is aware of an associated JobDetail.
7 JobDetail jobDetail = (JobDetail) trigger.getJobDataMap().remove("jobDetail");
8 if (triggerExists) {
9 if (jobDetail != null && !this.jobDetails.contains(jobDetail) && addJobToScheduler(jobDetail)) {
10 this.jobDetails.add(jobDetail);
11 }
12 try {
13 getScheduler().rescheduleJob(trigger.getKey(), trigger);
14 }
15 catch (ObjectAlreadyExistsException ex) {
16 if (logger.isDebugEnabled()) {
17 logger.debug("Unexpectedly encountered existing trigger on rescheduling, assumably due to "
18 "cluster race condition: " ex.getMessage() " - can safely be ignored");
19 }
20 }
21 }
22 else {
23 try {
24 if (jobDetail != null && !this.jobDetails.contains(jobDetail) &&
25 (this.overwriteExistingJobs || getScheduler().getJobDetail(jobDetail.getKey()) == null)) {
26 getScheduler().scheduleJob(jobDetail, trigger);
27 this.jobDetails.add(jobDetail);
28 }
29 else {
30 getScheduler().scheduleJob(trigger);
31 }
32 }
33 catch (ObjectAlreadyExistsException ex) {
34 if (logger.isDebugEnabled()) {
35 logger.debug("Unexpectedly encountered existing trigger on job scheduling, assumably due to "
36 "cluster race condition: " ex.getMessage() " - can safely be ignored");
37 }
38 if (this.overwriteExistingJobs) {
39 getScheduler().rescheduleJob(trigger.getKey(), trigger);
40 }
41 }
42 }
43 return true;
44}- 源码299行： addJobToScheduler(jobDetail) 一直会调用到 RAMJobStore 进行存放任务信息到 HashMap(100) 1public void storeJob(JobDetail newJob, 2 boolean replaceExisting) throws ObjectAlreadyExistsException { 3 JobWrapper jw = new JobWrapper((JobDetail)newJob.clone()); 4 boolean repl = false; 5 synchronized (lock) { 6 if (jobsByKey.get(jw.key) != null) { 7 if (!replaceExisting) { 8 throw new ObjectAlreadyExistsException(newJob); 9 } 10 repl = true; 11 } 12 if (!repl) { 13 // get job group 14 HashMap<JobKey, JobWrapper> grpMap = jobsByGroup.get(newJob.getKey().getGroup()); 15 if (grpMap == null) { 16 grpMap = new HashMap<JobKey, JobWrapper>(100); 17 jobsByGroup.put(newJob.getKey().getGroup(), grpMap); 18 } 19 // add to jobs by group 20 grpMap.put(newJob.getKey(), jw); 21 // add to jobs by FQN map 22 jobsByKey.put(jw.key, jw); 23 } else { 24 // update job detail 25 JobWrapper orig = jobsByKey.get(jw.key); 26 orig.jobDetail = jw.jobDetail; // already cloned 27 } 28 } 29}
- 初始化线程组；
SimpleThreadPool.initialize() & 这里的count是默认配置中的数量，可以更改
1 // create the worker threads and start them 2 Iterator<WorkerThread> workerThreads = createWorkerThreads(count).iterator(); 3 while(workerThreads.hasNext()) { 4 WorkerThread wt = workerThreads.next(); 5 wt.start(); 6 availWorkers.add(wt); 7 }
- prepareScheduler
- createScheduler
- schedulerFactory
- StdSchedulerFactory.getScheduler()
- getScheduler()->instantiate()
- 源码1323行： tp.initialize();
5. 启动定时任务
案例中使用硬编码方式调用 schedulerFactoryBean.start() 启动线程服务。线程的协作通过Object sigLock来实现，关于sigLock.wait()方法都在QuartzSchedulerThread的run方法里面，所以sigLock唤醒的是只有线程QuartzSchedulerThread。核心流程如下；
微信公众号：bugstack虫洞栈 & 调度启动流程
这个启动过程中，核心的代码类，如下；
- StdScheduler
- QuartzScheduler
- QuartzSchedulerThread
- ThreadPool
- RAMJobStore
- CronTriggerImpl
- JobRunShellFactory
代码语言：javascript复制QuartzScheduler.start() & 启动
1public void start() throws SchedulerException {
2
3 if (shuttingDown|| closed) {
4 throw new SchedulerException(
5 "The Scheduler cannot be restarted after shutdown() has been called.");
6 }
7
8 // QTZ-212 : calling new schedulerStarting() method on the listeners
9 // right after entering start()
10 notifySchedulerListenersStarting();
11
12 if (initialStart == null) {
13 initialStart = new Date();
14 this.resources.getJobStore().schedulerStarted();
15 startPlugins();
16 } else {
17 resources.getJobStore().schedulerResumed();
18 }
19
20 // 唤醒线程
21 schedThread.togglePause(false);
22
23 getLog().info(
24 "Scheduler " resources.getUniqueIdentifier() " started.");
25
26 notifySchedulerListenersStarted();
27}代码语言：javascript复制QuartzSchedulerThread.run() & 执行过程
1@Override
2public void run() {
3 int acquiresFailed = 0;
4
5 // 只有调用了halt()方法，才会退出这个死循环
6 while (!halted.get()) {
7 try {
8
9 // 一、如果是暂停状态，则循环超时等待1000毫秒
10
11 // wait a bit, if reading from job store is consistently failing (e.g. DB is down or restarting)..
12
13 // 阻塞直到有空闲的线程可用并返回可用的数量
14 int availThreadCount = qsRsrcs.getThreadPool().blockForAvailableThreads();
15 if(availThreadCount > 0) {
16
17 List<OperableTrigger> triggers;
18 long now = System.currentTimeMillis();
19 clearSignaledSchedulingChange();
20
21 try {
22 // 二、获取acquire状态的Trigger列表，也就是即将执行的任务
23 triggers = qsRsrcs.getJobStore().acquireNextTriggers(
24 now idleWaitTime, Math.min(availThreadCount, qsRsrcs.getMaxBat
25 acquiresFailed = 0;
26 if (log.isDebugEnabled())
27 log.debug("batch acquisition of " (triggers == null ? 0 : triggers
28 } catch(){//...}
29
30 if (triggers != null && !triggers.isEmpty()) {
31
32 // 三：获取List第一个Trigger的下次触发时刻
33 long triggerTime = triggers.get(0).getNextFireTime().getTime();
34
35 // 四：获取任务触发集合
36 List<TriggerFiredResult> res = qsRsrcs.getJobStore().triggersFired(triggers);
37
38 // 五：设置Triggers为'executing'状态
39 qsRsrcs.getJobStore().releaseAcquiredTrigger(triggers.get(i));
40
41 // 六：创建JobRunShell
42 qsRsrcs.getJobRunShellFactory().createJobRunShell(bndle);
43
44 // 七：执行Job
45 qsRsrcs.getThreadPool().runInThread(shell)
46
47 continue; // while (!halted)
48 }
49 } else { // if(availThreadCount > 0)
50 // should never happen, if threadPool.blockForAvailableThreads() follows con
51 continue; // while (!halted)
52 }
53
54
55 } catch(RuntimeException re) {
56 getLog().error("Runtime error occurred in main trigger firing loop.", re);
57 }
58 }
59
60 qs = null;
61 qsRsrcs = null;
62}- 源码391行： 创建JobRunShell，JobRunShell实例在initialize()方法就会把包含业务逻辑类的JobDetailImpl设置为它的成员属性，为后面执行业务逻辑代码做准备。执行业务逻辑代码在runInThread(shell)方法里面。 QuartzSchedulerThread.run() & 部分代码 1JobRunShell shell = null; 2try { 3 shell = qsRsrcs.getJobRunShellFactory().createJobRunShell(bndle); 4 shell.initialize(qs); 5} catch (SchedulerException se) { 6 qsRsrcs.getJobStore().triggeredJobComplete(triggers.get(i), bndle.getJobDetail(), CompletedExecutionInstruction.SET_ALL_JOB_TRIGGERS_ERROR); 7 continue; 8}
- 源码398行： qsRsrcs.getThreadPool().runInThread(shell) SimpleThreadPool.runInThread 1// 保存所有WorkerThread的集合 2private List<WorkerThread> workers; 3// 空闲的WorkerThread集合 4private LinkedList<WorkerThread> availWorkers = new LinkedList<WorkerThread>(); 5// 任务的WorkerThread集合 6private LinkedList<WorkerThread> busyWorkers = new LinkedList<WorkerThread>(); 7 8/** 9 * 维护workers、availWorkers和busyWorkers三个列表数据 10 * 有任务需要一个线程出来执行：availWorkers.removeFirst();busyWorkers.add() 11 * 然后调用WorkThread.run(runnable)方法 12 */ 13public boolean runInThread(Runnable runnable) { 14 if (runnable == null) { 15 return false; 16 }synchronized (nextRunnableLock) { 17 18 handoffPending = true; 19 20 // Wait until a worker thread is available 21 while ((availWorkers.size() < 1) && !isShutdown) { 22 try { 23 nextRunnableLock.wait(500); 24 } catch (InterruptedException ignore) { 25 } 26 } 27 28 if (!isShutdown) { 29 WorkerThread wt = (WorkerThread)availWorkers.removeFirst(); 30 busyWorkers.add(wt); 31 wt.run(runnable); 32 } else { 33 // If the thread pool is going down, execute the Runnable 34 // within a new additional worker thread (no thread from the pool). 35 36 WorkerThread wt = new WorkerThread(this, threadGroup, 37 "WorkerThread-LastJob", prio, isMakeThreadsDaemons(), runnable); 38 busyWorkers.add(wt); 39 workers.add(wt); 40 wt.start(); 41 } 42 nextRunnableLock.notifyAll(); 43 handoffPending = false; 44} 45 46return true; 47}
- 源码428行： WorkerThread ，是一个内部类，主要是赋值并唤醒lock对象的等待线程队列 WorkerThread.run(Runnable newRunnable) 1public void run(Runnable newRunnable) { 2 synchronized(lock) { 3 if(runnable != null) { 4 throw new IllegalStateException("Already running a Runnable!"); 5 } 6 runnable = newRunnable; 7 lock.notifyAll(); 8 } 9}
- 源码561行： WorkerThread 的run方法，方法执行lock.notifyAll()后，对应的WorkerThread就会来到run()方法。到这！接近曙光了！终于来到了执行业务的execute()方法的倒数第二步，runnable对象是一个JobRunShell对象，下面在看JobRunShell.run()方法。 WorkerThread.run() 1@Override 2public void run() { 3 boolean ran = false;while (run.get()) { 4 try { 5 synchronized(lock) { 6 while (runnable == null && run.get()) { 7 lock.wait(500); 8 } 9 if (runnable != null) { 10 ran = true; 11 // 启动真正执行的内容，runnable就是JobRunShell 12 runnable.run(); 13 } 14 } 15 } cache(){//...} 16} 17//if (log.isDebugEnabled()) 18try { 19 getLog().debug("WorkerThread is shut down."); 20} catch(Exception e) { 21 // ignore to help with a tomcat glitch 22} 23}
代码语言：javascript复制JobRunShell.run() & 从上面WorkerThread.run()，调用到这里执行
1public void run() {
2 qs.addInternalSchedulerListener(this);
3
4 try {
5 OperableTrigger trigger = (OperableTrigger) jec.getTrigger();
6 JobDetail jobDetail = jec.getJobDetail();
7
8 do {
9 // ...
10
11 long startTime = System.currentTimeMillis();
12 long endTime = startTime;
13
14 // execute the job
15 try {
16 log.debug("Calling execute on job " jobDetail.getKey());
17
18 // 执行业务代码，也就是我们的task
19 job.execute(jec);
20
21 endTime = System.currentTimeMillis();
22 } catch (JobExecutionException jee) {
23 endTime = System.currentTimeMillis();
24 jobExEx = jee;
25 getLog().info("Job " jobDetail.getKey()
26 " threw a JobExecutionException: ", jobExEx);
27 } catch (Throwable e) {
28 endTime = System.currentTimeMillis();
29 getLog().error("Job " jobDetail.getKey()
30 " threw an unhandled Exception: ", e);
31 SchedulerException se = new SchedulerException(
32 "Job threw an unhandled exception.", e);
33 qs.notifySchedulerListenersError("Job ("
34 jec.getJobDetail().getKey()
35 " threw an exception.", se);
36 jobExEx = new JobExecutionException(se, false);
37 }
38
39 jec.setJobRunTime(endTime - startTime);
40
41 // 其他代码
42 } while (true);
43
44 } finally {
45 qs.removeInternalSchedulerListener(this);
46 }
47}代码语言：javascript复制QuartzJobBean.execte() & 继续往下走
1public final void execute(JobExecutionContext context) throws JobExecutionException {
2 try {
3 BeanWrapper bw = PropertyAccessorFactory.forBeanPropertyAccess(this);
4 MutablePropertyValues pvs = new MutablePropertyValues();
5 pvs.addPropertyValues(context.getScheduler().getContext());
6 pvs.addPropertyValues(context.getMergedJobDataMap());
7 bw.setPropertyValues(pvs, true);
8 }
9 catch (SchedulerException ex) {
10 throw new JobExecutionException(ex);
11 }
12 executeInternal(context);
13}代码语言：javascript复制MethodInvokingJobDetailFactoryBean->MethodInvokingJob.executeInternal(JobExecutionContext context)
1protected void executeInternal(JobExecutionContext context) throws JobExecutionException {
2 try {
3 // 反射执行业务代码
4 context.setResult(this.methodInvoker.invoke());
5 }
6 catch (InvocationTargetException ex) {
7 if (ex.getTargetException() instanceof JobExecutionException) {
8 // -> JobExecutionException, to be logged at info level by Quartz
9 throw (JobExecutionException) ex.getTargetException();
10 }
11 else {
12 // -> "unhandled exception", to be logged at error level by Quartz
13 throw new JobMethodInvocationFailedException(this.methodInvoker, ex.getTargetException());
14 }
15 }
16 catch (Exception ex) {
17 // -> "unhandled exception", to be logged at error level by Quartz
18 throw new JobMethodInvocationFailedException(this.methodInvoker, ex);
19 }
20}五、综上总结
- quartz，即石英的意思，隐喻如石英钟般对时间的准确把握。
- 源码分析是一个很快乐的过程，这个快乐是分析完才能获得的快乐。纵横交互的背后是面向对象的高度解耦，对线程精彩的使用，将任务执行做成计划单，简直是一个超级棒的作品。
- 对于quartz.properties，简单场景下，开发者不用自定义配置，使用quartz默认配置即可，但在要求较高的使用场景中还是要自定义配置，比如通过org.quartz.threadPool.threadCount设置足够的线程数可提高多job场景下的运行性能。
- quartz 对任务处理高度解耦，job与trigger解藕，将任务本身和任务执行策略解藕，这样可以方便实现N个任务和M个执行策略自由组合。
- scheduler单独分离出来，相当于一个指挥官，可以从全局做调度，比如监听哪些trigger已经ready、分配线程等等。
- 外部链接：
- http://www.quartz-scheduler.org
- quartz-2.1.x/configuration
