分布式任务调度(Quartz/Elastic-Job)
Quartz
<!--引入quartz相关jar包-->
<dependency>
<groupId>org.quartz-scheduler</groupId>
<artifactId>quartz</artifactId>
<version>2.3.2</version>
</dependency>
package com.lagou;
import org.quartz.*;
import org.quartz.impl.StdSchedulerFactory;
/**
* @autor zixian
* @date 2021/6/22
*/
public class QuartzDemo {
public static void main(String[] args) throws SchedulerException {
//创建任务调度器
Scheduler scheduler = getScheduler();
//创建任务
JobDetail jobDetail = getJobDetail();
//创建任务触发器
Trigger trigger = getTrigger();
//使⽤调度器按照时间触发器执⾏这个任务
scheduler.scheduleJob(jobDetail, trigger);
scheduler.start();
}
/**
* 创建任务调度器
* @return
* @throws SchedulerException
*/
private static Scheduler getScheduler() throws SchedulerException {
SchedulerFactory factory = new StdSchedulerFactory();
Scheduler scheduler = factory.getScheduler();
return scheduler;
}
/**
* 创建一个任务
* @return
*/
private static JobDetail getJobDetail() {
JobDetail jobDetail = JobBuilder.newJob(MyJob.class).withIdentity("jobName").build();
return jobDetail;
}
/**
* 创建任务时间触发器
* cron表达式由七个位置组成,空格分隔
* 1、Seconds(秒) 0~59
* 2、Minutes(分) 0~59
* 3、Hours(⼩时) 0~23
* 4、Day of Month(天)1~31,注意有的⽉份不⾜31天
* 5、Month(⽉) 0~11,或者 JAN,FEB,MAR,APR,MAY,JUN,JUL,AUG,SEP,OCT,NOV,DEC
* 6、Day of Week(周) 1~7,1=SUN或者 SUN,MON,TUE,WEB,THU,FRI,SAT定义⼀个job,需实现Job接⼝
* 7、Year(年)1970~2099 可选项
* 示例:
* 0 0 11 * * ? 每天的11点触发执⾏⼀次
* 0 30 10 1 * ? 每⽉1号上午10点半触发执⾏⼀次
*/
private static Trigger getTrigger() {
Trigger trigger = TriggerBuilder.newTrigger()
.withIdentity("triggerName")
.startNow()
.withSchedule(CronScheduleBuilder.cronSchedule("0/2 * * * * ?"))
.build();
return trigger;
}
}
package com.lagou;
import org.quartz.Job;
import org.quartz.JobExecutionContext;
import org.quartz.JobExecutionException;
/**
* @autor zixian
* @date 2021/6/22
*/
public class MyJob implements Job {
@Override
public void execute(JobExecutionContext jobExecutionContext) throws JobExecutionException {
System.out.println("开始干活了。。。");
}
}
说明:如果想通过Quartz来实现分布式任务调度,则需要把Quartz部署成一个单独的任务调度服务器,然后调用对应服务器中的逻辑,比如调用集群中的restfull风格的接口,但是存在一个问题,调度服务器是单体的,存在一定的宕机风险
Elastic-Job
1、基本概念
elastic-job是当当网基于quartz进行的二次开发的分布式任务调度方案,有两个相对独立的子项目Elastic-Job-Lite和Elastic-Job-Cloud组成。本文主要分析 Elastic-Job-Lite,它定位为轻量级无中心化解决方案,使用jar包的形式提供分布式任务的协调服务,而Elastic-Job-Cloud子项目需要结合Mesos以及Docker在云环境下使用
2、elastic-job源码github下载地址: https://github.com/apache/shardingsphere-elasticjob
3、Elastic-Job-Lite应用
elastic-job依赖于Zookeeper进行分布式协调,需要安装Zookeeper,我这里安装的是zookeeper-3.4.10.tar.gz 版本,请自行安装。
<dependencies>
<!--引入quartz相关jar包-->
<dependency>
<groupId>org.quartz-scheduler</groupId>
<artifactId>quartz</artifactId>
<version>2.3.2</version>
</dependency>
<!--elastic-job-lite核心包-->
<dependency>
<groupId>com.dangdang</groupId>
<artifactId>elastic-job-lite-core</artifactId>
<version>2.1.5</version>
</dependency>
<dependency>
<groupId>mysql</groupId>
<artifactId>mysql-connector-java</artifactId>
<version>5.1.46</version>
</dependency>
</dependencies>
package com.lagou;
import java.sql.*;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
public class JdbcUtil {
//url
private static String url = "jdbc:mysql://10.60.131.17:3306/user?characterEncoding=utf8";
//user
private static String user = "root";
//password
private static String password = "123qweASD!@#";
//驱动程序类
private static String driver = "com.mysql.jdbc.Driver";
static {
try {
Class.forName(driver);
} catch (ClassNotFoundException e) {
e.printStackTrace();
}
}
public static Connection getConnection() {
try {
return DriverManager.getConnection(url, user, password);
} catch (SQLException e) {
e.printStackTrace();
}
return null;
}
public static void close(ResultSet rs, PreparedStatement ps, Connection con) {
if (rs != null) {
try {
rs.close();
} catch (SQLException e) {
e.printStackTrace();
} finally {
if (ps != null) {
try {
ps.close();
} catch (SQLException e) {
e.printStackTrace();
} finally {
if (con != null) {
try {
con.close();
} catch (SQLException e) {
e.printStackTrace();
}
}
}
}
}
}
}
/***
* DML操作(增删改)
* 1.获取连接数据库对象
* 2.预处理
* 3.执行更新操作
* @param sql
* @param obj
*/
//调用者只需传入一个sql语句,和一个Object数组。该数组存储的是SQL语句中的占位符
public static void executeUpdate(String sql,Object...obj) {
Connection con = getConnection();//调用getConnection()方法连接数据库
PreparedStatement ps = null;
try {
ps = con.prepareStatement(sql);//预处理
for (int i = 0; i < obj.length; i++) {//预处理声明占位符
ps.setObject(i + 1, obj[i]);
}
ps.executeUpdate();//执行更新操作
} catch (SQLException e) {
e.printStackTrace();
} finally {
close(null, ps, con);//调用close()方法关闭资源
}
}
/***
* DQL查询
* Result获取数据集
*
* @param sql
* @param obj
* @return
*/
public static List<Map<String,Object>> executeQuery(String sql, Object...obj) {
Connection con = getConnection();
ResultSet rs = null;
PreparedStatement ps = null;
try {
ps = con.prepareStatement(sql);
for (int i = 0; i < obj.length; i++) {
ps.setObject(i + 1, obj[i]);
}
rs = ps.executeQuery();
//new 一个空的list集合用来存放查询结果
List<Map<String, Object>> list = new ArrayList<>();
//获取结果集的列数
int count = rs.getMetaData().getColumnCount();
//对结果集遍历每一条数据是一个Map集合,列是k,值是v
while (rs.next()) {
//一个空的map集合,用来存放每一行数据
Map<String, Object> map = new HashMap<String, Object>();
for (int i = 0; i < count; i++) {
Object ob = rs.getObject(i + 1);//获取值
String key = rs.getMetaData().getColumnName(i + 1);//获取k即列名
map.put(key, ob);
}
list.add(map);
}
return list;
} catch (SQLException e) {
e.printStackTrace();
} finally {
close(rs, ps, con);
}
return null;
}
}
package com.lagou;
import com.dangdang.ddframe.job.api.ShardingContext;
import com.dangdang.ddframe.job.api.simple.SimpleJob;
import java.util.List;
import java.util.Map;
/**
* 数据归档job
* @autor zixian
* @date 2021/6/23
*/
public class ElasticMyJob implements SimpleJob {
@Override
public void execute(ShardingContext shardingContext) {
int shardingItem = shardingContext.getShardingItem();
Integer shardingParameter = Integer.valueOf(shardingContext.getShardingParameter());
System.out.println("======>>>threadId=" + Thread.currentThread().getId() + " 当前分片:" + shardingItem);
System.out.println("======>>>threadId=" + Thread.currentThread().getId() + " 当前分片参数:" + shardingParameter);
//查询出需要处理的数据
String sql = null;
//根据分片传入的参数执行不同的sql
if (shardingParameter.equals(1)) {
sql = "select * from resume where state='未处理' and id%2=1 limit 1";
} else {
sql = "select * from resume where state='未处理' and id%2=0 limit 1";
}
List<Map<String, Object>> dataList = JdbcUtil.executeQuery(sql);
if (dataList == null || dataList.size() == 0) {
System.out.println("处理完成");
}
Map<String, Object> map = dataList.get(0);
System.out.println("======>>>threadId=" + Thread.currentThread().getId() + " " + map.toString());
Long id = (Long) map.get("id");
String name = (String) map.get("name");
String sex = (String) map.get("sex");
String phone = (String) map.get("phone");
String address = (String) map.get("address");
String education = (String) map.get("education");
//处理数据
String updateStateSql = "update resume set state='已处理' where id=?";
JdbcUtil.executeUpdate(updateStateSql, id);
String insertResumeBakSql = "insert into resume_bak(name, sex, phone, address, education, state) values(?,?,?,?,?,?)";
JdbcUtil.executeUpdate(insertResumeBakSql, name, sex, phone, address, education, "已处理");
}
}
package com.lagou;
import com.dangdang.ddframe.job.config.JobCoreConfiguration;
import com.dangdang.ddframe.job.config.simple.SimpleJobConfiguration;
import com.dangdang.ddframe.job.lite.api.JobScheduler;
import com.dangdang.ddframe.job.lite.config.LiteJobConfiguration;
import com.dangdang.ddframe.job.reg.base.CoordinatorRegistryCenter;
import com.dangdang.ddframe.job.reg.zookeeper.ZookeeperConfiguration;
import com.dangdang.ddframe.job.reg.zookeeper.ZookeeperRegistryCenter;
/**
* @autor zixian
* @date 2021/6/22
*/
public class ElasticJobDemo {
public static void main(String[] args) {
//初始化zk
ZookeeperConfiguration zkConfig = new ZookeeperConfiguration("10.60.131.17:2181", "data-sharding-job");
CoordinatorRegistryCenter coordinatorRegistryCenter = new ZookeeperRegistryCenter(zkConfig);
coordinatorRegistryCenter.init();
//配置任务,将任务分成两个分片,根据id进行分片,奇数id一个分片任务,偶数id一个分片任务
JobCoreConfiguration job = JobCoreConfiguration.newBuilder("sharding-job-1", "0/2 * * * * ?", 2)
.shardingItemParameters("0=1,1=2")//1-奇数 2-偶数
.build();
SimpleJobConfiguration simpleJobConfiguration = new SimpleJobConfiguration(job, ElasticMyJob.class.getName());
//启动任务
new JobScheduler(coordinatorRegistryCenter, LiteJobConfiguration.newBuilder(simpleJobConfiguration).build()).init();
}
}
测试结果:
1、启动了一个节点,两个分片都在改节点中运行
2、启动了两个节点,每个节点中运行一个分片
4、总结
当我们需要执行一个非常耗时的job时,我们可以使用Elastic-Job的横向扩展能力,将任务job分成若干个分片,每个分片执行什么逻辑由我们自己定制,每个分片交由一个机器实例来执行(一个机器实例也可以处理多个分片),当我们扩容或者缩容后,定时任务下次执行时,分片算法会重新执行,如果机器挂掉到只剩一个节点,所有分片都会指向剩下的一个节点,这也是Elastic-Job高可用的体现。