Hystrix-request collapsing(请求合并)
介绍:
Hystrix的请求合并就是把重复的请求批量的用一个HystrixCommand命令去执行，以减少通信消耗和线程数的占用。Hystrix的请求合并用到了HystrixCollapser这个抽象类，它在HystrixCommand之前前放置一个合并处理器，将处于一个很短的时间窗(默认10ms)内对同一依赖服务的多个请求进行整合并以批量方式发起请求的功能（服务提供方也需要提供相应的匹狼实现接口）。下面我们通过一个例子来看看怎么使用。
示例:
这个示例是基于spring cloud的，对此不熟悉的可以参考这里了解。
1.首先我们需要一个EurekaServer来作为注册中心。这个没什么特别的说明，可以参考代码示例中的 eureka-server工程
2.新建一个服务提供者工程eureka-server-service
2.1 新建一个User.java。这个model必须要有一个无参的默认构造器，否则后面的实验会报错
package org.hope.model; public class User { //必须要有一个无参的构造器 public User(){} public User(Long id, String name) { this.id = id; this.name = name; } private Long id; private String name; setters()&getters(); }
2.2 新建一个提供服务的controller
package org.hope.web; import org.hope.model.User; import org.springframework.web.bind.annotation.PathVariable; import org.springframework.web.bind.annotation.RequestMapping; import org.springframework.web.bind.annotation.RequestMethod; import org.springframework.web.bind.annotation.RestController; import java.util.ArrayList; import java.util.List; @RestController public class UserBatchController { @RequestMapping(value = "/users", method = RequestMethod.GET) public List<User> batchUser(String ids) { System.out.println("ids===:" + ids); List<User> lists = new ArrayList<User>(); lists.add(new User(1l, "小明")); lists.add(new User(2l, "小红")); lists.add(new User(3l, "小张")); lists.add(new User(4l, "小王")); lists.add(new User(5l, "小李")); return lists; } @RequestMapping(value = "/users/{id}", method = RequestMethod.GET) public User singleUser(@PathVariable("id") String id) { User user = new User(100L, "大王"); return user; } }
2.3 springboot的main函数
package org.hope; import org.springframework.beans.factory.annotation.Value; import org.springframework.boot.SpringApplication; import org.springframework.boot.autoconfigure.SpringBootApplication; import org.springframework.cloud.client.discovery.EnableDiscoveryClient; import org.springframework.cloud.netflix.eureka.EnableEurekaClient; import org.springframework.web.bind.annotation.RequestMapping; import org.springframework.web.bind.annotation.RequestParam; import org.springframework.web.bind.annotation.RestController; @SpringBootApplication @EnableEurekaClient @RestController public class ServiceApplication { public static void main(String[] args) { SpringApplication.run(ServiceApplication.class, args); } }
2.4 配置文件application.yml
eureka:
client:
serviceUrl:
#注册中心的地址
defaultZone: http://localhost:8761/eureka/
server:
#当前服务端口号
port: 8762
spring:
application:
#当前应用名称
name: service-batch
3.新建一个服务消费者工程ribbon-hystrix，在这个工程里我们测试hystrix批量服务调用的请求合并功能
3.1新建一个model User.java。这个model一定要有无参的构造器
package org.hope.lee.model; public class User { //一定要有无参的构造器 public User(){} public User(Long id, String name) { this.id = id; this.name = name; } private Long id; private String name; getters()&setters(); }
3.2 service负责调用服务
package org.hope.lee.service; import org.hope.lee.model.User; import java.util.List; public interface UserService { public User find(Long id); public List<User> findAll(List<Long> ids); }
package org.hope.lee.service; import org.apache.commons.lang.ArrayUtils; import org.apache.commons.lang.StringUtils; import org.hope.lee.model.User; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.stereotype.Service; import org.springframework.web.client.RestTemplate; import java.util.Arrays; import java.util.List; @Service("userService") public class UserServiceImpl implements UserService{ @Autowired private RestTemplate restTemplate; @Override public User find(Long id) { return restTemplate.getForObject("http://localhost:8762/users/{1}",User.class, id); } @Override public List<User> findAll(List<Long> ids) { System.out.println("finaAll request:---------" + ids + "Thread.currentThread().getName():-------" + Thread.currentThread().getName()); User[] users = restTemplate.getForObject("http://localhost:8762/users?ids={1}", User[].class, StringUtils.join(ids, ",")); return Arrays.asList(users); } }
3.3 HystrixCommand命令执行请求
package org.hope.lee.command; import com.netflix.hystrix.HystrixCommand; import com.netflix.hystrix.HystrixCommandGroupKey; import com.netflix.hystrix.HystrixCommandKey; import org.hope.lee.model.User; import org.hope.lee.service.UserService; import java.util.ArrayList; import java.util.List; public class UserBatchCommand extends HystrixCommand<List<User>> { UserService userService; List<Long> userIds; public UserBatchCommand(UserService userService, List<Long> userIds) { super(Setter.withGroupKey(HystrixCommandGroupKey.Factory.asKey("ExampleGroup")). andCommandKey(HystrixCommandKey.Factory.asKey("GetValueForKey"))); this.userService = userService; this.userIds = userIds; } @Override protected List<User> run() throws Exception { return userService.findAll(userIds); } @Override protected List<User> getFallback() { List<User> users = new ArrayList<User>(); users.add(new User(99L, "失败者")); return new ArrayList<User>(users); } }
3.4 HystrixCollapser命令来做请求合并
package org.hope.lee.command; import com.netflix.hystrix.HystrixCollapser; import com.netflix.hystrix.HystrixCollapserKey; import com.netflix.hystrix.HystrixCollapserProperties; import com.netflix.hystrix.HystrixCommand; import org.hope.lee.model.User; import org.hope.lee.service.UserService; import java.util.ArrayList; import java.util.Collection; import java.util.List; import java.util.stream.Collectors;
/**
* Created by lisen on 2017/12/27.
* 通过看HystrixCollapser类的源码:
* public abstract class HystrixCollapser<BatchReturnType, ResponseType, RequestArgumentType>
* 我们可以知道List<User>表示:合并后批量请求的返回类型
* User表示：单个请求返回的类型
* Long表示:请求参数类型
*/
public class UserCollapseCommand extends HystrixCollapser<List<User>, User, Long> { private UserService userService; private Long userId; public UserCollapseCommand(UserService userService, Long userId) { super(Setter.withCollapserKey(HystrixCollapserKey.Factory.asKey("userCollapseCommand")). andCollapserPropertiesDefaults(HystrixCollapserProperties.Setter().withTimerDelayInMilliseconds(100))); this.userService = userService; this.userId = userId; } @Override public Long getRequestArgument() { return userId; } /** * * @param collapsedRequests 保存了延迟时间窗中收集到的所有获取单个User的请求。通过获取这些请求的参数来组织 * 我们准备的批量请求命令UserBatchCommand实例 * @return */ @Override protected HystrixCommand<List<User>> createCommand(Collection<CollapsedRequest<User, Long>> collapsedRequests) { List<Long> userIds = new ArrayList<>(collapsedRequests.size()); userIds.addAll(collapsedRequests.stream().map(CollapsedRequest::getArgument).collect(Collectors.toList())); return new UserBatchCommand(userService, userIds); } /** * 在批量请求命令UserBatchCommand实例被触发执行完成后，该方法开始执行， * 在这里我们通过批量结果batchResponse对象，为collapsedRequests中每个合并前的单个请求设置返回结果。 * 来完成批量结果到单个请求结果的转换 * @param batchResponse 保存了createCommand中组织的批量请求命令的返回结果 * @param collapsedRequests 代表了每个合并的请求 */ @Override protected void mapResponseToRequests(List<User> batchResponse, Collection<CollapsedRequest<User, Long>> collapsedRequests) { System.out.println("mapResponseToRequests========>"); int count = 0; for(CollapsedRequest<User, Long> collapsedRequest : collapsedRequests) { User user = batchResponse.get(count++); collapsedRequest.setResponse(user); } } }
3.5写一个controller来辅助测试
package org.hope.lee.web; import com.netflix.hystrix.strategy.concurrency.HystrixRequestContext; import org.hope.lee.command.UserCollapseCommand; import org.hope.lee.model.User; import org.hope.lee.service.UserService; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.web.bind.annotation.RequestMapping; import org.springframework.web.bind.annotation.RequestMethod; import org.springframework.web.bind.annotation.RestController; import java.util.concurrent.Future; @RestController public class CollapseCommandController { @Autowired private UserService userService; @RequestMapping(value = "/collapse", method = RequestMethod.GET) public void requestCollapse() { HystrixRequestContext context = HystrixRequestContext.initializeContext(); try { Future<User> f1 = new UserCollapseCommand(userService, 1L).queue(); Future<User> f2 = new UserCollapseCommand(userService, 2L).queue(); Future<User> f3 = new UserCollapseCommand(userService, 3L).queue(); Thread.sleep(3000); Future<User> f4 = new UserCollapseCommand(userService, 4L).queue(); Future<User> f5 = new UserCollapseCommand(userService, 5L).queue(); User u1 = f1.get(); User u2 = f2.get(); User u3 = f3.get(); User u4 = f4.get(); User u5 = f5.get(); System.out.println(u1.getName()); System.out.println(u2.getName()); System.out.println(u3.getName()); System.out.println(u4.getName()); System.out.println(u5.getName()); } catch (Exception e) { e.printStackTrace(); }finally { context.close(); } } }
3.6 spring boot main方法
package org.hope.lee; import org.springframework.boot.SpringApplication; import org.springframework.boot.autoconfigure.SpringBootApplication; import org.springframework.cloud.client.discovery.EnableDiscoveryClient; import org.springframework.cloud.client.loadbalancer.LoadBalanced; import org.springframework.cloud.netflix.hystrix.EnableHystrix; import org.springframework.cloud.netflix.hystrix.dashboard.EnableHystrixDashboard; import org.springframework.context.annotation.Bean; import org.springframework.web.client.RestTemplate; @SpringBootApplication @EnableDiscoveryClient public class ServiceRibbonApplication { public static void main(String[] args) { SpringApplication.run(ServiceRibbonApplication.class, args); } @Bean RestTemplate restTemplate() { return new RestTemplate(); } }
3.7.配置文件applicationi.properties
eureka.client.serviceUrl.defaultZone=http://localhost:8761/eureka/
server.port=8763
spring.application.name=batch-customer
测试：
1.运行eureka-server启动注册中心
2.运行eureka-server-service启动服务提供者
3.运行ribbon-hystrix启动服务消费者
4.在浏览器输入: http://localhost:8763/collapse
输出结果:从结果中我们看到前3次请求合并为一个请求，后面2次请求合并为了一个请求
遇到的问题:
model的实体类一定要定义一个无参数的默认构造器。否则就会报错。
使用注解实现请求合并器
package org.hope.lee.service; import com.netflix.hystrix.contrib.javanica.annotation.HystrixCollapser; import com.netflix.hystrix.contrib.javanica.annotation.HystrixCommand; import com.netflix.hystrix.contrib.javanica.annotation.HystrixProperty; import org.apache.commons.lang.StringUtils; import org.hope.lee.model.User; import org.springframework.beans.factory.annotation.Autowired; import org.springframework.stereotype.Service; import org.springframework.web.client.RestTemplate; import java.util.Arrays; import java.util.List; import java.util.concurrent.Future; @Service("peopleService") public class PeopleServiceImpl implements PeopleService { @Autowired private RestTemplate restTemplate; @HystrixCollapser(batchMethod = "findAll", collapserProperties = {@HystrixProperty(name = "timerDelayInMilliseconds", value = "100")}) public Future<User> find(Long id) { throw new RuntimeException("This method body should not be executed"); } @HystrixCommand public List<User> findAll(List<Long> ids) { System.out.println("Annotation---------" + ids + "Thread.currentThread().getName():" + Thread.currentThread().getName()); User[] users = restTemplate.getForObject("http://localhost:8762/users?ids={1}", User[].class, StringUtils.join(ids, ",")); return Arrays.asList(users); } }
https://gitee.com/huayicompany/Hystrix-learn/tree/master/hystrix-request-collapsing
参考：
[1]博客，https://segmentfault.com/a/1190000011468804，Spring Cloud中Hystrix的请求合并
[2]官网，https://github.com/Netflix/Hystrix/wiki/How-To-Use#Collapsing， Request Collapsing
[3]《SpringCloud微服务实战》，电子工业出版社，翟永超
