如何优化Springboot线程池并发处理数据方式

这篇文章给大家分享的是有关如何优化Springboot线程池并发处理数据方式的内容。小编觉得挺实用的，因此分享给大家做个参考，一起跟随小编过来看看吧。

第一步：首先配置线程基本参数

可以放在application.propertes文件种也可以放在自己新建的config/文件目录下，注意：但是需要使用@PropertySource把配置文件进行加载。

# 异步线程配置
# 配置核心线程数
async.executor.thread.core_pool_size = 8
# 配置最大线程数
async.executor.thread.max_pool_size = 20
# 配置队列大小
async.executor.thread.queue_capacity = 99999
# 配置线程池中的线程的名称前缀
async.executor.thread.name.prefix = async-service-

第二步：让Spring Boot加载

用来定义如何创建一个ThreadPoolTaskExecutor，要使用@Configuration和@EnableAsync这两个注解，表示这是个配置类，并且是线程池的配置类

@Slf4j
@EnableAsync
@Configuration
public class RCExecutorConfig {
 
  @Value("${async.executor.thread.core_pool_size}")
     private int corePoolSize;
     @Value("${async.executor.thread.max_pool_size}")
     private int maxPoolSize;
     @Value("${async.executor.thread.queue_capacity}")
     private int queueCapacity;
     @Value("${async.executor.thread.name.prefix}")
     private String namePrefix;
    
     @Bean(name = "asyncServiceExecutor")
     public Executor asyncServiceExecutor() {
         log.info("start asyncServiceExecutor");
         ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
         //ThreadPoolTaskExecutor executor = new VisiableThreadPoolTaskExecutor();
         //配置核心线程数
         executor.setCorePoolSize(corePoolSize);
         //配置最大线程数
         executor.setMaxPoolSize(maxPoolSize);
         //配置队列大小
         executor.setQueueCapacity(queueCapacity);
         //配置线程池中的线程的名称前缀
         executor.setThreadNamePrefix(namePrefix);
 
         // rejection-policy：当pool已经达到max size的时候，如何处理新任务
         // CALLER_RUNS：不在新线程中执行任务，而是有调用者所在的线程来执行
         executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
         //执行初始化
         executor.initialize();
         return executor;
     }
}

第三步：创建一个service接口

是异步线程的接口，便于测试

public interface AsyncService {
 
 /**
     *  执行异步任务
     *  可以根据需求，自己加参数拟定
     */
    void executeAsync();
}

第四步：编写现实类

将Service层的服务异步化，在executeAsync()方法上增加注解@Async("asyncServiceExecutor")，asyncServiceExecutor方法是前面RCExecutorConfig.java中的方法名，表明executeAsync方法进入的线程池是asyncServiceExecutor方法创建的。

测试方面这里我加入了一个定时任务，使用的是corn表达式。（不懂得同学可以网上了解一下）

@Slf4j
@Service
public class AsyncServiceImpl implements AsyncService {
 
 @Override
 @Scheduled(cron = " */1 * * * * ? ")
 @Async("asyncServiceExecutor")
 public void executeAsync() {
  log.info("start executeAsync");
        System.out.println("异步线程执行批量插入等耗时任务");
        log.info("end executeAsync");  
 }  
}

第五步：测试结果如下

10:32:15.004 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:15.004 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:16.003 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:16.004 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:17.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:17.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:18.002 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:18.003 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:19.002 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:19.003 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:20.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:20.002 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:21.002 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:21.002 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:22.004 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:22.005 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:23.001 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:23.003 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:24.003 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:24.003 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:25.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:25.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:26.002 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:26.002 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:27.002 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:27.003 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:28.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:28.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:29.001 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:29.002 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:30.001 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:30.001 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:32:31.001 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:32:31.001 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync

还没完：

通过以上日志可以发现，[async-service-]是有多个线程的，显然已经在我们配置的线程池中执行了，表明每次请求都快速响应了，而耗时的操作都留给线程池中的线程去异步执行；

还有另一个问提就是，虽然已经用上了线程池，但是依然不清楚线程池当时的情况，有多少线程在执行，多少在队列中等待呢？于是这里我创建了一个ThreadPoolTaskExecutor的子类，可以把每次提交线程的时候都会将当前线程池的运行状况打印出来

import lombok.extern.slf4j.Slf4j;
 
@Slf4j
public class VisiableThreadPoolTaskExecutor extends  ThreadPoolTaskExecutor {
 
    /**
  * 
  */
 private static final long serialVersionUID = -3518460523928455463L; 
 private void showThreadPoolInfo(String prefix) {
        ThreadPoolExecutor threadPoolExecutor = getThreadPoolExecutor(); 
        if (null == threadPoolExecutor) {
            return;
        }
 
        log.info("{}, {},taskCount [{}], completedTaskCount [{}], activeCount [{}], queueSize [{}]",
                this.getThreadNamePrefix(),
                prefix,
                threadPoolExecutor.getTaskCount(),
                threadPoolExecutor.getCompletedTaskCount(),
                threadPoolExecutor.getActiveCount(),
                threadPoolExecutor.getQueue().size());
    }
 
    @Override
    public void execute(Runnable task) {
        showThreadPoolInfo("1. do execute");
        super.execute(task);
    }
 
    @Override
    public void execute(Runnable task, long startTimeout) {
        showThreadPoolInfo("2. do execute");
        super.execute(task, startTimeout);
    }
 
    @Override
    public Future<?> submit(Runnable task) {
        showThreadPoolInfo("1. do submit");
        return super.submit(task);
    }
 
    @Override
    public <T> Future<T> submit(Callable<T> task) {
        showThreadPoolInfo("2. do submit");
        return super.submit(task);
    }
 
    @Override
    public ListenableFuture<?> submitListenable(Runnable task) {
        showThreadPoolInfo("1. do submitListenable");
        return super.submitListenable(task);
    }
 
    @Override
    public <T> ListenableFuture<T> submitListenable(Callable<T> task) {
        showThreadPoolInfo("2. do submitListenable");
        return super.submitListenable(task);
    } 
}

其次：修改RCExecutorConfig.java的asyncServiceExecutor方法，

将ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor()改为ThreadPoolTaskExecutor executor = new VisiableThreadPoolTaskExecutor()

 @Bean(name = "asyncServiceExecutor")
     public Executor asyncServiceExecutor() {
         log.info("start asyncServiceExecutor");
         //ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
         ThreadPoolTaskExecutor executor = new VisiableThreadPoolTaskExecutor();
         //配置核心线程数
         executor.setCorePoolSize(corePoolSize);
         //配置最大线程数
         executor.setMaxPoolSize(maxPoolSize);
         //配置队列大小
         executor.setQueueCapacity(queueCapacity);
         //配置线程池中的线程的名称前缀
         executor.setThreadNamePrefix(namePrefix); 
         // rejection-policy：当pool已经达到max size的时候，如何处理新任务
         // CALLER_RUNS：不在新线程中执行任务，而是有调用者所在的线程来执行
         executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
         //执行初始化
         executor.initialize();
         return executor;
     }

测试结果如下：

异步线程执行批量插入等耗时任务
10:41:35.003 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:36.001 [sheduled-pool-1-thread-1] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [5], completedTaskCount [5], activeCount [0], queueSize [0]
10:41:36.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:36.002 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:37.001 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [6], completedTaskCount [6], activeCount [0], queueSize [0]
10:41:37.001 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:37.002 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:38.002 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [7], completedTaskCount [7], activeCount [0], queueSize [0]
10:41:38.002 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:38.002 [async-service-8] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:39.001 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [8], completedTaskCount [8], activeCount [0], queueSize [0]
10:41:39.001 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:39.002 [async-service-1] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:40.003 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [9], completedTaskCount [9], activeCount [0], queueSize [0]
10:41:40.003 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:40.003 [async-service-2] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:41.001 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [10], completedTaskCount [10], activeCount [0], queueSize [0]
10:41:41.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:41.001 [async-service-3] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:42.000 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [11], completedTaskCount [11], activeCount [0], queueSize [0]
10:41:42.000 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:42.000 [async-service-4] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:43.001 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [12], completedTaskCount [12], activeCount [0], queueSize [0]
10:41:43.002 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:43.003 [async-service-5] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:44.001 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [13], completedTaskCount [13], activeCount [0], queueSize [0]
10:41:44.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync
异步线程执行批量插入等耗时任务
10:41:44.001 [async-service-6] INFO c.a.a.service.impl.AsyncServiceImpl - end executeAsync
10:41:45.000 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [14], completedTaskCount [14], activeCount [0], queueSize [0]
10:41:45.001 [async-service-7] INFO c.a.a.service.impl.AsyncServiceImpl - start executeAsync

解释：这里意思提交了14个任务，处理了14个任务，对列中还剩0个任务

10:41:45.000 [sheduled-pool-1-thread-7] INFO c.a.a.e.t.VisiableThreadPoolTaskExecutor - async-service-, 2. do submit,taskCount [14], completedTaskCount [14], activeCount [0], queueSize [0]

感谢各位的阅读！关于“如何优化Springboot线程池并发处理数据方式”这篇文章就分享到这里了，希望以上内容可以对大家有一定的帮助，让大家可以学到更多知识，如果觉得文章不错，可以把它分享出去让更多的人看到吧！

亿速云「云服务器」，即开即用、新一代英特尔至强铂金CPU、三副本存储NVMe SSD云盘，价格低至29元/月。点击查看>>