Quartz Cluster Enhancement_00.How to use?

Quartz Cluster Enhancement_00.How to use?

Reproduced with attribution/funnyzpc/p/18540378

open source address/funnyzpc/quartz

Basic structure of a table

• 

In general the configuration and development of the tasks basically follows the basic relationship of the table in the figure above, except for the app as well asnode The configuration needs to be done manually, except for theapp up tonode is automatically generated when the executable is started.app as well asnode data ~

Rear Tube Configuration

Let's start with the basic page of the back tube~
on account ofapp together withnode It's a one-to-many relationship, so here it goes under one page:

• 

  Here's what needs to be saidapptogether withnodeGenerally there is no need to add a new one, if in special cases please refer to the following figure:

app additions

• 

node added

• 

on account ofnodeMust be associated with an existingappYou can only add new entries, which are available in theapp listcenter

In addition, it should be noted that.

• If the executing side can't get the hosthost IPas well asHost Namewill randomly generate a same-namedhost IPas well asHost NameAt this point, it makes no sense to add them manually on the admin side

  removing

• 

• To delete an application you must first delete the node associated with the application (node), a node being deleted prevents the node's corresponding execution side from executing its tasks, and deleting the application also

• Deleting an application or a node does not change the state of tasks and executables, nor does it delete tasks and executables. Executables without a node are not executed and are periodically cleaned up.

  Enable/Disable

• 

Enable and disable only operate on nodes or applications, disabling a node prevents all tasks under the node from executing, disabling an application prevents all nodes associated with the application from executing tasks, and this operation does not change the tasks or executables~.

Take a look at the node task and execution configuration again.

• 

Task/execution configuration is the main task of the management side, the execution configuration using the associated task configuration (PID) associated with the corresponding task (job), the execution item (execute) is not independent of the existence!

New Task Configuration

• 

  The application name/scheduling name is the automatically or manually configuredApplication Information
  The task state can only have initialization at configuration time (INIT)/normal execution (EXECUTING), if you don't want to execute it immediately, choose Initialize (INIT)

  Added Execute Config-CRON time task

• 

  The task type can only be a simple task (SIMPLE) or a task with a time item of expression (CRON), with both types of execution configured (execute) Filling in the fields will make a difference
  CRON mission(used form a nominal expression)CRON expressionis required, and the default time zone at this stage isAsia/Shanghai , which will be changed to get the default from the system.
  Starting timeGenerally do not fill in the default is -1, the new submission is based on the current time to replenish the
  end timeis also non-required, the end time is also -1 by default, the end time if -1 will be added to the last execution time after the last execution of the task

  Added Execute Configuration-SIMPLE Time Tasks

• 

  The circled fields in the chart are required, and it should be noted that if theImplementation end timetogether withNumber of executionsAll set, specific tasks will be executed according to the smallest limit for the actual execution, such as set the end of a long time but the number of times to execute only a few times, then the final probability will only be the number of times to execute the limit for execution

  Additionally, for execution configurations, when execution is complete, the correspondingExecute ConfigurationDelete only, do not modify or stop, has been completed for such operations is meaningless, rather than adding a new execution configuration

Console Development Configuration and Integration

Here is an example of springboot only:

• Add a dependency, if you have a maven private service, it is recommended that you put it in the private service.

    <dependency>
        <groupId></groupId>
        <artifactId>quartz-client</artifactId>
        <version>2.3.2</version>
        <!-- This is a local introduction，Suggested to be put into private service-->
        <scope>system</scope>
        <systemPath>${}/src/main/resources/lib/quartz-client-2.3.</systemPath>
    </dependency>

• Startup classes need to exclude auto-assembly

// This line is the kicker!
@SpringBootApplication(exclude = {})
public class MeeAdminApplication {
/**
* Logging
*/
private static final Logger LOG= ();

public static void main(String[] args)throws Exception {
ConfigurableApplicationContext application = (, args);
Environment env = ();
String ip = ().getHostAddress();

String path = ("-path"); String
("\n\t----------------------------------------------------------\n\t" +
"Application MeeAdminApplication is running!\n\t" +
"Local: \t\thttp://localhost:" + port + path + "/\n\t" +
"External: \thttp://" + ip + ":" + port + path + "/\n\t" +
"----------------------------------------------------------");
}

}

• An instance needs to be configured to use the

@Service
public final class QrtzJobServiceImpl implements QrtzJobService {

    /**
     * log (computing)
     */
    private static final Logger LOG = ();
    
    /**
     * quartztimed taskapi
     */
    private final Scheduler scheduler;

    public QrtzJobServiceImpl(DataSource dataSource) {
         = new StdScheduler(dataSource);
    }
}

• Calling the sdk

    @Override
    public MeeResult<Integer> updateJobState(String job_id,String state) {
        Object[] result = (job_id,state);
        int updateCount = (int)result[0];
        if(updateCount>0){
            return (updateCount);
        }else{
            return ((String)result[1]);
        }
    }

Scheduler provides a wide variety of api's, note the difference between some of the interfaces:

in the event thatmanagement sidetogether withexecutive sideAll-in-one There is no need to introduce a client dependency (quartz-client), and there is no need to exclude auto-assembly from the startup class (QuartzAutoConfiguration), and you don't need to pass in a database using the sdk constructor, just this.

    @Autowired
    private Scheduler scheduler;

Execution-side development configuration and integration

• Introducing dependencies while excluding native Quartz

        <dependency>
            <groupId></groupId>
            <artifactId>spring-boot-starter-quartz</artifactId>
            <version>${}</version>
            <exclusions>
                <exclusion>
                    <groupId>-scheduler</groupId>
                    <artifactId>quartz</artifactId>
                </exclusion>
            </exclusions>
        </dependency>
        <dependency>
            <groupId></groupId>
            <artifactId>quartz-core</artifactId>
            <version>2.3.2</version>
            <!-- This is a local introduction，Suggested to be put into private service-->
            <scope>system</scope>
            <systemPath>${}/src/main/resources/lib/quartz-core-2.3.</systemPath>
        </dependency>

• Adding Dependencies

### ----------- quartz ------------------
-store-type=jdbc
=
=6000
= =true
=true
# Table name prefix
=QRTZ_
=${}
#=
==
# Thread count configuration
=10
=5
# Thread inheritance context class loader for initialization threads
=true
# Whether to enable pessimistic lock to control trigger concurrency in the cluster
=true

Inside the configuration item, pay attention to the number of threads, if you are using theMeeThreadPool imitatethreadCountis the maximum number of threads, the number of core threadsthreadCount-2 The minimum number of CPU cores is 2, depending on the actual number of CPU cores and whether it is IO-intensive or CPU-intensive.
secondly, it is important to note thattablePrefix If the table name has been changed, the changed table name will be used.prefix (linguistics)Just configure it.

• Define a task
  • If you are using thespringofferedQuartzJobBeanto develop:

    import ;
    import ;
    import ;
    import ;
    import org.;
    import org.;
    import ;
    import ;
    
    import ;
    
    
      public class ATestJob extends QuartzJobBean {
    
      private static final Logger log = ();
    
        @Override
        protected void executeInternal(JobExecutionContext context) throws JobExecutionException {
            try {
                ("===>ATestJob::executeInternal {}-{} : {}-{}<===" ,(),(),(),());
            } catch (Exception e) {
                throw new JobExecutionException(e);
            }
        }
    
    }
  

  • If you are using theQuartzofferedJob interfaceto develop, also available:

    import ;
    import ;
    import ;
    import ;
    import org.;
    import org.;
    
    import ;
    
    public class Job01TestService  implements Job {
    
    private static final Logger LOGGER = ();
    
    @Override
    public void execute(JobExecutionContext context) throws JobExecutionException {
        ("=>>{}-{}.{}-{}",(),(),(),());
    }
  }
  
  

Both of the above are possible, and it is important to note that whether you inherit theQuartzJobBean or an implementation of ``Job, there is no need to make the class famous as aspring beanClasses (@Service or @Component),QuartzThe internal self-creates the task class asspring bean ~

Development Notes

• utilizationquartz-client Added tasks are generally most likely to be added at the end of the5 seconds. is executed after that, because task polling is5 seconds. poll
• The executing side performs an exception (Quartznon-operational) tasks at the latest in the15SResume task execution afterward, since cluster/default fire handling is15 seconds. poll
• Added tasks are noted if they are not executed first Configuration item is configured or not, this configuration item corresponds to theapp tabulatedapplication field
• Actual tasks with logs showing task delays are recommended for troubleshooting.hosting resourcesWhether it is occupied or not, orNumber of threadsIs the configuration reasonable