前言
上一篇文章 Spring整合Quartz分布式调度 介绍了Quartz通过数据库的方式来实现分布式调度,通过使用数据库来存储trigger,job等信息,可以在停服重启的时候重新加载上次trigger的状态,保证了完整性;另一方面通过数据库来实现锁机制来实现分布式调度;Quartz默认提供了11张表,本文将对这几张表做简要的分析。
1.qrtz_blob_triggers 2.qrtz_cron_triggers 3.qrtz_simple_triggers 4.qrtz_simprop_triggers 5.qrtz_fired_triggers 6.qrtz_triggers 7.qrtz_job_details 8.qrtz_calendars 9.qrtz_paused_trigger_grps 10.qrtz_scheduler_state 11.qrtz_locks
共11张表,前6张都是关于各种triggers的信息,后面包括job,悲观锁,调度状态等信息;相关表操作在类StdJDBCDelegate中,相关sql语句在StdJDBCConstants中;
1.qrtz_blob_triggers
自定义的triggers使用blog类型进行存储,非自定义的triggers不会存放在此表中,Quartz提供的triggers包括:CronTrigger,CalendarIntervalTrigger,
DailyTimeIntervalTrigger以及SimpleTrigger,这几个trigger信息会保存在后面的几张表中;
2.qrtz_cron_triggers
存储CronTrigger,这也是我们使用最多的触发器,在配置文件中做如下配置,即可在qrtz_cron_triggers生成记录:
<bean id="firstCronTrigger" class="org.springframework.scheduling.quartz.CronTriggerFactoryBean"> <property name="jobDetail" ref="firstTask" /> <property name="cronExpression" value="0/6 * * ? * *" /> <property name="group" value="firstCronGroup"></property> </bean> <bean id="firstTask" class="org.springframework.scheduling.quartz.JobDetailFactoryBean"> <property name="jobClass" value="zh.maven.SQuartz.task.FirstTask" /> <property name="jobDataMap"> <map> <entry key="firstService" value-ref="firstService" /> </map> </property> </bean> <bean id="firstService" class="zh.maven.SQuartz.service.FirstService"></bean>
表达式指定了每隔6秒执行一次,然后指定了要执行的task,task指定了要执行的业务,运行之后可以查看数据表:
mysql> select * from qrtz_cron_triggers; +-------------+------------------+----------------+-----------------+---------------+ | SCHED_NAME | TRIGGER_NAME | TRIGGER_GROUP | CRON_EXPRESSION | TIME_ZONE_ID | +-------------+------------------+----------------+-----------------+---------------+ | myScheduler | firstCronTrigger | firstCronGroup | 0/6 * * ? * * | Asia/Shanghai | +-------------+------------------+----------------+-----------------+---------------+
myScheduler是在定义SchedulerFactoryBean时指定的名称,其他字段都可以在上面的配置中找到;
3.qrtz_simple_triggers
存储SimpleTrigger,在配置文件中做如下配置,即可在qrtz_simple_triggers生成记录:
<bean id="firstSimpleTrigger" class="org.springframework.scheduling.quartz.SimpleTriggerFactoryBean"> <property name="jobDetail" ref="firstSimpleTask" /> <property name="startDelay" value="1000" /> <property name="repeatInterval" value="2000" /> <property name="repeatCount" value="5"></property> <property name="group" value="firstSimpleGroup"></property> </bean> <bean id="firstSimpleTask" class="org.springframework.scheduling.quartz.JobDetailFactoryBean"> <property name="jobClass" value="zh.maven.SQuartz.task.SimpleFirstTask" /> <property name="jobDataMap"> <map> <entry key="firstService" value-ref="simpleFirstService" /> </map> </property> </bean> <bean id="simpleFirstService" class="zh.maven.SQuartz.service.SimpleFirstService"></bean>
指定了开始延迟时间,重复间隔时间已经重复的次数限制,查看表如下:
mysql> select * from qrtz_simple_triggers; +-------------+--------------------+------------------+--------------+-----------------+-----------------+ | SCHED_NAME | TRIGGER_NAME | TRIGGER_GROUP | REPEAT_COUNT | REPEAT_INTERVAL | TIMES_TRIGGERED | +-------------+--------------------+------------------+--------------+-----------------+-----------------+ | myScheduler | firstSimpleTrigger | firstSimpleGroup | 5 | 2000 | 1 | +-------------+--------------------+------------------+--------------+-----------------+-----------------+
TIMES_TRIGGERED用来记录执行了多少次了,此值被定义在SimpleTriggerImpl中,每次执行+1,这里定义的REPEAT_COUNT=5,实际情况会执行6次,具体可以查看SimpleTriggerImpl源码:
public Date getFireTimeAfter(Date afterTime) { if (complete) { return null; } if ((timesTriggered > repeatCount) && (repeatCount != REPEAT_INDEFINITELY)) { return null; } ...... }
timesTriggered默认值为0,当timesTriggered > repeatCount停止trigger,所以会执行6次,当执行完毕之后此记录会被删除;
4.qrtz_simprop_triggers
存储CalendarIntervalTrigger和DailyTimeIntervalTrigger两种类型的触发器,使用CalendarIntervalTrigger做如下配置:
<bean id="firstCalendarTrigger" class="org.quartz.impl.triggers.CalendarIntervalTriggerImpl"> <property name="jobDataMap"> <map> <entry key="jobDetail" value-ref="firstCalendarTask"></entry> </map> </property> <property name="key" ref="calendarTriggerKey"></property> <property name="repeatInterval" value="1" /> <property name="group" value="firstCalendarGroup"></property> </bean> <bean id="firstCalendarTask" class="org.springframework.scheduling.quartz.JobDetailFactoryBean"> <property name="jobClass" value="zh.maven.SQuartz.task.CalendarFirstTask" /> <property name="jobDataMap"> <map> <entry key="firstService" value-ref="calendarFirstService" /> </map> </property> </bean> <bean id="calendarFirstService" class="zh.maven.SQuartz.service.CalendarFirstService"></bean>
CalendarIntervalTrigger没有对应的FactoryBean,直接设置实现类CalendarIntervalTriggerImpl;指定的重复周期是1,默认单位是天,也就是每天执行一次,查看表如下:
mysql> select * from qrtz_simprop_triggers; +-------------+--------------------+--------------------+------------+---------------+------------+------------+------------+-------------+-------------+------------+------------+-------------+-------------+ | SCHED_NAME | TRIGGER_NAME | TRIGGER_GROUP | STR_PROP_1 | STR_PROP_2 | STR_PROP_3 | INT_PROP_1 | INT_PROP_2 | LONG_PROP_1 | LONG_PROP_2 | DEC_PROP_1 | DEC_PROP_2 | BOOL_PROP_1 | BOOL_PROP_2 | +-------------+--------------------+--------------------+------------+---------------+------------+------------+------------+-------------+-------------+------------+------------+-------------+-------------+ | myScheduler | calendarTriggerKey | firstCalendarGroup | DAY | Asia/Shanghai | NULL | 1 | 1 | 0 | 0 | NULL | NULL | 0 | 0 | +-------------+--------------------+--------------------+------------+---------------+------------+------------+------------+-------------+-------------+------------+------------+-------------+-------------+
提供了3个string类型的参数,2个int类型的参数,2个long类型的参数,2个decimal类型的参数以及2个boolean类型的参数;具体每个参数是什么含义,根据不同的trigger类型存放各自的参数;
5.qrtz_fired_triggers
存储已经触发的trigger相关信息,trigger随着时间的推移状态发生变化,直到最后trigger执行完成,从表中被删除;已SimpleTrigger为例重复3次执行,查询表:
mysql> select * from qrtz_fired_triggers; +-------------+----------------------------------------+--------------------+------------------+---------------------------+---------------+---------------+----------+-----------+-----------------+-----------+------------------+-------------------+ | SCHED_NAME | ENTRY_ID | TRIGGER_NAME | TRIGGER_GROUP | INSTANCE_NAME | FIRED_TIME | SCHED_TIME | PRIORITY | STATE | JOB_NAME | JOB_GROUP | IS_NONCONCURRENT | REQUESTS_RECOVERY | +-------------+----------------------------------------+--------------------+------------------+---------------------------+---------------+---------------+----------+-----------+-----------------+-----------+------------------+-------------------+ | myScheduler | NJD9YZGJ2-PC15241041777351524104177723 | firstSimpleTrigger | firstSimpleGroup | NJD9YZGJ2-PC1524104177735 | 1524104178499 | 1524104178472 | 0 | EXECUTING | firstSimpleTask | DEFAULT | 0 | 0 | | myScheduler | NJD9YZGJ2-PC15241041777351524104177724 | firstSimpleTrigger | firstSimpleGroup | NJD9YZGJ2-PC1524104177735 | 1524104180477 | 1524104180472 | 0 | EXECUTING | firstSimpleTask | DEFAULT | 0 | 0 | | myScheduler | NJD9YZGJ2-PC15241041777351524104177725 | firstSimpleTrigger | firstSimpleGroup | NJD9YZGJ2-PC1524104177735 | 1524104180563 | 1524104182472 | 0 | ACQUIRED | NULL | NULL | 0 | 0 | +-------------+----------------------------------------+--------------------+------------------+---------------------------+---------------+---------------+----------+-----------+-----------------+-----------+------------------+-------------------+
相同的trigger和task,每触发一次都会创建一个实例;从刚被创建的ACQUIRED状态,到EXECUTING状态,最后执行完从数据库中删除;
6.qrtz_triggers
存储定义的trigger,以上定义的三个triggers为例,分别是:firstSimpleTrigger,firstCalendarTrigger和firstCronTrigger,运行之后查看数据库:
mysql> select * from qrtz_triggers; +-------------+--------------------+--------------------+-------------------+-----------+-------------+----------------+----------------+----------+---------------+--------------+---------------+----------+---------------+---------------+----------+ | SCHED_NAME | TRIGGER_NAME | TRIGGER_GROUP | JOB_NAME | JOB_GROUP | DESCRIPTION | NEXT_FIRE_TIME | PREV_FIRE_TIME | PRIORITY | TRIGGER_STATE | TRIGGER_TYPE | START_TIME | END_TIME | CALENDAR_NAME | MISFIRE_INSTR | JOB_DATA | +-------------+--------------------+--------------------+-------------------+-----------+-------------+----------------+----------------+----------+---------------+--------------+---------------+----------+---------------+---------------+----------+ | myScheduler | calendarTriggerKey | firstCalendarGroup | firstCalendarTask | DEFAULT | NULL | 1524203884719 | 1524117484719 | 5 | WAITING | CAL_INT | 1524117484719 | 0 | NULL | 0 | | | myScheduler | firstCronTrigger | firstCronGroup | firstTask | DEFAULT | NULL | 1524117492000 | 1524117486000 | 0 | ACQUIRED | CRON | 1524117483000 | 0 | firstCalendar | 0 | | | myScheduler | firstSimpleTrigger | firstSimpleGroup | firstSimpleTask | DEFAULT | NULL | -1 | 1524117488436 | 0 | COMPLETE | SIMPLE | 1524117484436 | 0 | NULL | 0 | | +-------------+--------------------+--------------------+-------------------+-----------+-------------+----------------+----------------+----------+---------------+--------------+---------------+----------+---------------+---------------+----------+
和qrtz_fired_triggers存放的不一样,不管trigger触发了多少次都只有一条记录,TRIGGER_STATE用来标识当前trigger的状态;firstCalendarTask每天执行一次,执行完之后一直是WAITING状态;firstCronTrigger每6秒执行一次状态是ACQUIRED状态;firstSimpleTrigger重复执行6次后状态为COMPLETE,并且会被删除;
7.qrtz_job_details
存储jobDetails信息,相关信息在定义的时候指定,如上面定义的JobDetailFactoryBean,查询数据库:
mysql> select * from qrtz_job_details; +-------------+-------------------+-----------+-------------+-----------------------------------------+------------+------------------+----------------+-------------------+----------+ | SCHED_NAME | JOB_NAME | JOB_GROUP | DESCRIPTION | JOB_CLASS_NAME | IS_DURABLE | IS_NONCONCURRENT | IS_UPDATE_DATA | REQUESTS_RECOVERY | JOB_DATA | +-------------+-------------------+-----------+-------------+-----------------------------------------+------------+------------------+----------------+-------------------+----------+ | myScheduler | firstCalendarTask | DEFAULT | NULL | zh.maven.SQuartz.task.CalendarFirstTask | 0 | 0 | 0 | 0 | | | myScheduler | firstSimpleTask | DEFAULT | NULL | zh.maven.SQuartz.task.SimpleFirstTask | 0 | 0 | 0 | 0 | | | myScheduler | firstTask | DEFAULT | NULL | zh.maven.SQuartz.task.FirstTask | 0 | 0 | 0 | 0 | | +-------------+-------------------+-----------+-------------+-----------------------------------------+------------+------------------+----------------+-------------------+----------+
JOB_DATA存放的就是定义task时指定的jobDataMap属性,所以此属性需要实现Serializable接口,方便持久化到数据库;
8.qrtz_calendars
Quartz为我们提供了日历的功能,可以自己定义一个时间段,可以控制触发器在这个时间段内触发或者不触发;现在提供6种类型:AnnualCalendar,CronCalendar,DailyCalendar,
HolidayCalendar,MonthlyCalendar,WeeklyCalendar;以下使用CronCalendar为例:
<bean id="firstCalendar" class="org.quartz.impl.calendar.CronCalendar"> <constructor-arg value="0/5 * * ? * *"></constructor-arg> </bean> <bean id="firstCronTrigger" class="org.springframework.scheduling.quartz.CronTriggerFactoryBean"> <property name="jobDetail" ref="firstTask" /> <property name="cronExpression" value="0/6 * * ? * *" /> <property name="group" value="firstCronGroup"></property> <property name="calendarName" value="firstCalendar"></property> </bean> <bean id="scheduler" class="org.springframework.scheduling.quartz.SchedulerFactoryBean"> <property name="schedulerName" value="myScheduler"></property> <property name="dataSource" ref="dataSource" /> <property name="configLocation" value="classpath:quartz.properties" /> <property name="triggers"> <list> <ref bean="firstCronTrigger" /> </list> </property> <property name="calendars"> <map> <entry key="firstCalendar" value-ref="firstCalendar"></entry> </map> </property> </bean>
定义了一个排除每隔5秒的CronCalendar,然后在firstCronTrigger中指定了calendarName,并且需要在SchedulerFactoryBean中定义calendars;
因为firstCronTrigger每6秒执行一次,而CronCalendar排除每隔5秒,所以会出现firstCronTrigger在第5次触发的时候需要等待12秒,结果如下:
20180419 15:09:06---start FirstService 20180419 15:09:08---end FirstService 20180419 15:09:12---start FirstService 20180419 15:09:14---end FirstService 20180419 15:09:18---start FirstService 20180419 15:09:20---end FirstService 20180419 15:09:24---start FirstService 20180419 15:09:26---end FirstService 20180419 15:09:36---start FirstService 20180419 15:09:38---end FirstService
查询保存在数据中的CronCalendar:
mysql> select * from qrtz_calendars; +-------------+---------------+----------+ | SCHED_NAME | CALENDAR_NAME | CALENDAR | +-------------+---------------+----------+ | myScheduler | firstCalendar | | +-------------+---------------+----------+
CALENDAR存放的是CronCalendar序列化之后的数据;
9.qrtz_paused_trigger_grps
存放暂停掉的触发器,测试手动暂停firstCronTrigger,代码如下:
public class App { public static void main(String[] args) { final AbstractApplicationContext context = new ClassPathXmlApplicationContext("quartz.xml"); final StdScheduler scheduler = (StdScheduler) context.getBean("scheduler"); try { Thread.sleep(4000); scheduler.pauseTriggers(GroupMatcher.triggerGroupEquals("firstCronGroup")); } catch (Exception e) { e.printStackTrace(); } } }
启动之后延迟4秒后暂停firstCronTrigger,这里传递的参数group,然后查看数据库:
mysql> select * from qrtz_paused_trigger_grps; +-------------+----------------+ | SCHED_NAME | TRIGGER_GROUP | +-------------+----------------+ | myScheduler | firstCronGroup | +-------------+----------------+
因为已经入库,所以重启之后firstCronGroup还是处于暂停状态,firstCronTrigger不会运行;
10.qrtz_scheduler_state
存储所有节点的scheduler,会定期检查scheduler是否失效,启动多个scheduler,查询数据库:
mysql> select * from qrtz_scheduler_state; +-------------+---------------------------+-------------------+------------------+ | SCHED_NAME | INSTANCE_NAME | LAST_CHECKIN_TIME | CHECKIN_INTERVAL | +-------------+---------------------------+-------------------+------------------+ | myScheduler | NJD9YZGJ2-PC1524209095408 | 1524209113973 | 1000 | | myScheduler | NJD9YZGJ2-PC1524209097649 | 1524209113918 | 1000 | +-------------+---------------------------+-------------------+------------------+
记录了最后最新的检查时间,在quartz.properties中设置了CHECKIN_INTERVAL为1000,也就是每秒检查一次;
11.qrtz_locks
Quartz提供的锁表,为多个节点调度提供分布式锁,实现分布式调度,默认有2个锁:
mysql> select * from qrtz_locks; +-------------+----------------+ | SCHED_NAME | LOCK_NAME | +-------------+----------------+ | myScheduler | STATE_ACCESS | | myScheduler | TRIGGER_ACCESS | +-------------+----------------+
STATE_ACCESS主要用在scheduler定期检查是否失效的时候,保证只有一个节点去处理已经失效的scheduler;
TRIGGER_ACCESS主要用在TRIGGER被调度的时候,保证只有一个节点去执行调度;
总结
本文对这11张表做了简要的分析,介绍了每张表具体是用来存储什么的,并且给了简单的实例;其实如果要实现一个trigger的管理系统,其实也就是对这几张表的维护。