基于springcloud异步线程池、高并发请求feign的解决方案

 更新时间:2021年2月25日 15:00  
这篇文章主要介绍了基于springcloud异步线程池、高并发请求feign的解决方案,具有很好的参考价值,希望对大家有所帮助。一起跟随小编过来看看吧

ScenTaskTestApplication.java

package com.test;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.cloud.openfeign.EnableFeignClients;
/**
* @author scen
* @version 2018年9月27日 上午11:51:04
*/
@EnableFeignClients
@SpringBootApplication
public class ScenTaskTestApplication {
 public static void main(String[] args) {
  SpringApplication.run(ScenTaskTestApplication.class, args);
 }
}

application.properties

spring.application.name=scen-task-test
server.port=9009
feign.hystrix.enabled=true
#熔断器失败的个数==进入熔断器的请求达到1000时服务降级(之后的请求直接进入熔断器)
hystrix.command.default.circuitBreaker.requestVolumeThreshold=1000
#回退最大线程数
hystrix.command.default.fallback.isolation.semaphore.maxConcurrentRequests=50
#核心线程池数量
hystrix.threadpool.default.coreSize=130
#请求处理的超时时间
hystrix.command.default.execution.isolation.thread.timeoutInMilliseconds=100000
ribbon.ReadTimeout=120000
#请求连接的超时时间
ribbon.ConnectTimeout=130000
eureka.instance.instance-id=${spring.application.name}:${spring.application.instance_id:${server.port}}
eureka.instance.preferIpAddress=true
eureka.client.service-url.defaultZone=http://127.0.0.1:9000/eureka
logging.level.com.test.user.service=debug
logging.level.org.springframework.boot=debug
logging.level.custom=info

AsyncConfig.java

package com.test;
import org.springframework.context.annotation.Configuration;
import org.springframework.scheduling.annotation.AsyncConfigurer;
import org.springframework.scheduling.annotation.EnableAsync;
import org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor;
import java.util.concurrent.Executor;
/**
 * springboot异步线程池配置
 * @author Scen
 * @date 2018/11/7 18:28
 */
@Configuration
@EnableAsync
public class AsyncConfig implements AsyncConfigurer {
 
 
 @Override
 public Executor getAsyncExecutor() {
  //定义线程池
  ThreadPoolTaskExecutor taskExecutor = new ThreadPoolTaskExecutor();
  //核心线程数
  taskExecutor.setCorePoolSize(20);
  //线程池最大线程数
  taskExecutor.setMaxPoolSize(100);
  //线程队列最大线程数
  taskExecutor.setQueueCapacity(10);
  //初始化
  taskExecutor.initialize();
  return taskExecutor;
 }
}

DoTaskClass.java

package com.test;
import com.test.pojo.User;
import com.test.pojo.UserEducation;
import com.test.user.service.UserService;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.scheduling.annotation.Async;
import org.springframework.stereotype.Component;
import java.util.List;
/**
 * 任务类 定义异步工作任务
 * @author Scen
 * @date 2018/11/7 18:40
 */
@Component
public class DoTaskClass { 
 /**
  * 一个feign的客户端
  */
 private final UserService userService;
 
 @Autowired
 public DoTaskClass(UserService userService) {
  this.userService = userService;
 }
 
 /**
  * 核心任务
  *
  * @param uid
  */
 @Async
 public void dotask(String uid) {
  /**
   * 模拟复杂工作业务(109个线程同时通过feign请求微服务提供者)
   */
  {
   List<UserEducation> userEducationByUid = userService.findUserEducationByUid(uid);
   List<String> blackList = userService.getBlackList();
   String userSkilled = userService.getUserSkilled(uid);
   String userFollow = userService.getUserFollow(uid);
   User userById = userService.getUserById(uid);
   List<String> followList = userService.getFollowList(uid);
   int userActivityScore = userService.getUserActivityScore(uid);
  }
//  打印线程名称分辨是否为多线程操作
  System.out.println(Thread.currentThread().getName() + "===任务" + uid + "执行完成===");
 }
}

TestController.java

package com.test;
import com.test.pojo.User;
import com.test.user.service.UserService;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;
import java.util.List;
/**
 * 测试案例
 * @author Scen
 * @date 2018/11/7 18:10
 */
@RestController
public class TestController {
 
 /**
  * 此处仅用此feign客户端请求微服务获取核心工作所需参数
  */
 private final UserService userService;
 
 /**
  * 核心工作异步算法
  */
 private final DoTaskClass doTaskClass;
 
 @Autowired
 public TestController(DoTaskClass doTaskClass, UserService userService) {
  this.doTaskClass = doTaskClass;
  this.userService = userService;
 } 
 
 /**
  * 手动触发工作
  * @throws InterruptedException
  */
 @RequestMapping("/test")
 public void task() throws InterruptedException {
  /*
   取到1000个要执行任务的必备参数
   */
  List<User> userList = userService.findAllLite(1, 1000);
  for (int i = 0; i < userList.size(); i++) {
   try {
//    异步线程开始工作
    doTaskClass.dotask(userList.get(i).getId());
   } catch (Exception e) {
    /*
     若并发线程数达到MaxPoolSize+QueueCapacity=110(参考AsyncConfig配置)会进入catch代码块
     i--休眠3秒后重试(尝试进入线程队列:当且仅当109个线程有一个或多个线程完成异步任务时重试成功)
     */
    i--;
    Thread.sleep(3000*3);
   }
   System.out.println(i);
  }
 } 
}

相关线程池、超时时间等数量和大小按实际业务配置

补充:[email protected]问题

@FeignClient可以作为Http代理访问其他微服务节点,[email protected]ion请求方式,以达到让http请求走Http线程池的目的。

[email protected],在hystrix dashboard上可以监控到 @FeignClient 中接口调用情况和 @FeignClient 中httpclient中线程池使用状况。

下面是demo的示例:

[email protected]

@FeignClient(value="service-A", fallback=ServiceClientHystrix.class)
public interface ServiceClient { 
 @RequestMapping(method = RequestMethod.GET, value = "/add/{id}")
 String add(@PathVariable("id") Integer id);
}

2、ServiceClientHystrix.java

@Component
public class ServiceClientHystrix implements ServiceClient{
 @Override
 public String add(Integer id) {
  return "add value from ServiceClientHystrix";
 }
}

[email protected]

集成后,Http线程池配置如下:

hystrix.threadpool.服务实例ID.参数

例如设置httpclient的线程池最大线程数量

hystrix.threadpool.service-A.coreSize=20//默认是hystrix.threadpool.default.coreSize = 10
hystrix.threadpool.service-A.maximumSize=20//默认是hystrix.threadpool.default.maximumSize = 10

启动服务后用测试用例连续调用接口测试,用hystrix dashboard

监控得到下图监控效果:

去掉hystrix.threadpool.服务实例ID.参数配置后,再次用测试用例调用接口得到监控如下图:

PoolSize的大小取决于hystrix.threadpool.服务实例ID.coreSize大小设置

以上为个人经验,希望能给大家一个参考,也希望大家多多支持猪先飞。如有错误或未考虑完全的地方,望不吝赐教。