Spring线程池ThreadPoolTaskExecutor配置及详情

Posted 2020-09-01 低调的小黑

Spring线程池ThreadPoolTaskExecutor配置及详情

1. ThreadPoolTaskExecutor配置

 1 <!-- spring thread pool executor -->           
 2     <bean id="taskExecutor" class="org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor">
 3         <!-- 线程池维护线程的最少数量 -->
 4         <property name="corePoolSize" value="5" />
 5         <!-- 允许的空闲时间 -->
 6         <property name="keepAliveSeconds" value="200" />
 7         <!-- 线程池维护线程的最大数量 -->
 8         <property name="maxPoolSize" value="10" />
 9         <!-- 缓存队列 -->
10         <property name="queueCapacity" value="20" />
11         <!-- 对拒绝task的处理策略 -->
12         <property name="rejectedExecutionHandler">
13             <bean class="java.util.concurrent.ThreadPoolExecutor$CallerRunsPolicy" />
14         </property>
15     </bean>

属性字段说明

corePoolSize：线程池维护线程的最少数量

keepAliveSeconds：允许的空闲时间

maxPoolSize：线程池维护线程的最大数量

queueCapacity：缓存队列

rejectedExecutionHandler：对拒绝task的处理策略

2. execute(Runable)方法执行过程

如果此时线程池中的数量小于corePoolSize，即使线程池中的线程都处于空闲状态，也要创建新的线程来处理被添加的任务。

如果此时线程池中的数量等于 corePoolSize，但是缓冲队列 workQueue未满，那么任务被放入缓冲队列。

如果此时线程池中的数量大于corePoolSize，缓冲队列workQueue满，并且线程池中的数量小于maxPoolSize，建新的线程来处理被添加的任务。

如果此时线程池中的数量大于corePoolSize，缓冲队列workQueue满，并且线程池中的数量等于maxPoolSize，那么通过handler所指定的策略来处理此任务。也就是：处理任务的优先级为：核心线程corePoolSize、任务队列workQueue、最大线程 maximumPoolSize，如果三者都满了，使用handler处理被拒绝的任务。

当线程池中的线程数量大于corePoolSize时，如果某线程空闲时间超过keepAliveTime，线程将被终止。这样，线程池可以动态的调整池中的线程数。

3. 示例代码

Junit Test

 1 @RunWith(SpringJUnit4ClassRunner.class)
 2 @ContextConfiguration(classes = { MultiThreadConfig.class })
 3 public class MultiThreadTest {
 4 
 5     @Autowired
 6     private ThreadPoolTaskExecutor taskExecutor;
 7 
 8     @Autowired
 9     private MultiThreadProcessService multiThreadProcessService;
10     
11     @Test
12     public void test() {
13         
14         int n = 20;
15         for (int i = 0; i < n; i++) {
16             taskExecutor.execute(new MultiThreadDemo(multiThreadProcessService));
17             System.out.println("int i is " + i + ", now threadpool active threads totalnum is " + taskExecutor.getActiveCount());
18         }
19         
20         try {
21             System.in.read();
22         } catch (IOException e) {
23             throw new RuntimeException(e);
24         }
25     }
26 }

MultiThreadDemo

 1 /**
 2  * 多线程并发处理demo
 3  * @author daniel.zhao
 4  *
 5  */
 6 public class MultiThreadDemo implements Runnable {
 7 
 8     private MultiThreadProcessService multiThreadProcessService;
 9     
10     public MultiThreadDemo() {
11     }
12     
13     public MultiThreadDemo(MultiThreadProcessService multiThreadProcessService) {
14         this.multiThreadProcessService = multiThreadProcessService;
15     }
16     
17     @Override
18     public void run() {
19         multiThreadProcessService.processSomething();
20     }
21 
22 }

MultiThreadProcessService

 1 @Service
 2 public class MultiThreadProcessService {
 3 
 4     public static final Logger logger = Logger.getLogger(MultiThreadProcessService.class);
 5     
 6     /**
 7      * 默认处理流程耗时1000ms
 8      */
 9     public void processSomething() {
10         logger.debug("MultiThreadProcessService-processSomething" + Thread.currentThread() + "......start");
11         try {
12             Thread.sleep(1000);
13         } catch (InterruptedException e) {
14             throw new RuntimeException(e);
15         }
16         logger.debug("MultiThreadProcessService-processSomething" + Thread.currentThread() + "......end");
17     }
18 }

MultiThreadConfig

1 @Configuration
2 @ComponentScan(basePackages = { "com.xxx.multithread" })
3 @ImportResource(value = { "classpath:config/application-task.xml" })
4 @EnableScheduling
5 public class MultiThreadConfig {
6 }