基于springcloud異步線程池、高并發(fā)請求feign的解決方案

更新時間：2021年02月25日 10:38:55 作者：Scenoob

這篇文章主要介紹了基于springcloud異步線程池、高并發(fā)請求feign的解決方案，具有很好的參考價值，希望對大家有所幫助。一起跟隨小編過來看看吧

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
#熔斷器失敗的個數(shù)==進入熔斷器的請求達到1000時服務(wù)降級（之后的請求直接進入熔斷器）
hystrix.command.default.circuitBreaker.requestVolumeThreshold=1000
#回退最大線程數(shù)
hystrix.command.default.fallback.isolation.semaphore.maxConcurrentRequests=50
#核心線程池數(shù)量
hystrix.threadpool.default.coreSize=130
#請求處理的超時時間
hystrix.command.default.execution.isolation.thread.timeoutInMilliseconds=100000
ribbon.ReadTimeout=120000
#請求連接的超時時間
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異步線程池配置
 * @author Scen
 * @date 2018/11/7 18:28
 */
@Configuration
@EnableAsync
public class AsyncConfig implements AsyncConfigurer {
 
 
 @Override
 public Executor getAsyncExecutor() {
  //定義線程池
  ThreadPoolTaskExecutor taskExecutor = new ThreadPoolTaskExecutor();
  //核心線程數(shù)
  taskExecutor.setCorePoolSize(20);
  //線程池最大線程數(shù)
  taskExecutor.setMaxPoolSize(100);
  //線程隊列最大線程數(shù)
  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;
/**
 * 任務(wù)類 定義異步工作任務(wù)
 * @author Scen
 * @date 2018/11/7 18:40
 */
@Component
public class DoTaskClass { 
 /**
  * 一個feign的客戶端
  */
 private final UserService userService;
 
 @Autowired
 public DoTaskClass(UserService userService) {
  this.userService = userService;
 }
 
 /**
  * 核心任務(wù)
  *
  * @param uid
  */
 @Async
 public void dotask(String uid) {
  /**
   * 模擬復(fù)雜工作業(yè)務(wù)（109個線程同時通過feign請求微服務(wù)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);
  }
//  打印線程名稱分辨是否為多線程操作
  System.out.println(Thread.currentThread().getName() + "===任務(wù)" + uid + "執(zhí)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;
/**
 * 測試案例
 * @author Scen
 * @date 2018/11/7 18:10
 */
@RestController
public class TestController {
 
 /**
  * 此處僅用此feign客戶端請求微服務(wù)獲取核心工作所需參數(shù)
  */
 private final UserService userService;
 
 /**
  * 核心工作異步算法
  */
 private final DoTaskClass doTaskClass;
 
 @Autowired
 public TestController(DoTaskClass doTaskClass, UserService userService) {
  this.doTaskClass = doTaskClass;
  this.userService = userService;
 } 
 
 /**
  * 手動觸發(fā)工作
  * @throws InterruptedException
  */
 @RequestMapping("/test")
 public void task() throws InterruptedException {
  /*
   取到1000個要執(zhí)行任務(wù)的必備參數(shù)
   */
  List<User> userList = userService.findAllLite(1, 1000);
  for (int i = 0; i < userList.size(); i++) {
   try {
//    異步線程開始工作
    doTaskClass.dotask(userList.get(i).getId());
   } catch (Exception e) {
    /*
     若并發(fā)線程數(shù)達到MaxPoolSize+QueueCapacity=110（參考AsyncConfig配置）會進入catch代碼塊
     i--休眠3秒后重試（嘗試進入線程隊列：當(dāng)且僅當(dāng)109個線程有一個或多個線程完成異步任務(wù)時重試成功）
     */
    i--;
    Thread.sleep(3000*3);
   }
   System.out.println(i);
  }
 } 
}

相關(guān)線程池、超時時間等數(shù)量和大小按實際業(yè)務(wù)配置

補充：SpringCloud關(guān)于@FeignClient和Hystrix集成對http線程池監(jiān)控問題

@FeignClient可以作為Http代理訪問其他微服務(wù)節(jié)點，可以用apache的httpclient替換@FeignClient原生的URLConnection請求方式，以達到讓http請求走Http線程池的目的。

而@FeignClient和hystrix集成之后，在hystrix dashboard上可以監(jiān)控到 @FeignClient 中接口調(diào)用情況和 @FeignClient 中httpclient中線程池使用狀況。

下面是demo的示例：

1、@FeignClient的接口代碼如下：

@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";
 }
}

3、關(guān)于@FeignClient和hystrix

集成后，Http線程池配置如下：

hystrix.threadpool.服務(wù)實例ID.參數(shù)

例如設(shè)置httpclient的線程池最大線程數(shù)量

hystrix.threadpool.service-A.coreSize=20//默認是hystrix.threadpool.default.coreSize = 10
hystrix.threadpool.service-A.maximumSize=20//默認是hystrix.threadpool.default.maximumSize = 10

啟動服務(wù)后用測試用例連續(xù)調(diào)用接口測試，用hystrix dashboard

監(jiān)控得到下圖監(jiān)控效果：