zuul-redislimiter-spring-boot基于Zuul网关Filter的分布式限流

儒雅程序员  

zuul-redislimiter-spring-boot

基于Zuul的限流器

 

快速开始

克隆, 编译，安装

git clone https://github.com/tangaiyun/zuul-redislimiter-spring-boot.git
cd zuulredislimiter/zuul-redislimiter-spring-boot-starter
mvn clean install

新建一个Zuul项目

参考项目 "zuulredislimiter/zuulapp"创建一个zuul网关项目，在pom.xml文件里添加必须的依赖

package com.tay.zuulapp;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.cloud.netflix.zuul.EnableZuulProxy;

@SpringBootApplication
@EnableZuulProxy
public class ZuulappApplication {
 public static void main(String[] args) {
 SpringApplication.run(ZuulappApplication.class, args);
 }
}

 <dependencies>
   <dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-netflix-zuul</artifactId>
   </dependency>
   <dependency>
    <groupId>org.springframework.cloud</groupId>
    <artifactId>spring-cloud-starter-netflix-eureka-client</artifactId>
   </dependency>
   <dependency>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-test</artifactId>
    <scope>test</scope>
   </dependency>
   <dependency>
    <groupId>com.tay</groupId>
    <artifactId>zuul-redislimiter-spring-boot-starter</artifactId>
    <version>0.0.1-SNAPSHOT</version>
   </dependency>
 </dependencies>

配置

修改 resources/application.yml 加入以下配置

server:
 port: 8000 
spring:
 application:
 name: service-zuul

zuul:
 routes:
 service1: /s1/**
 service2: /s2/**
 redis-limiter:
 redis-host: 127.0.0.1 #Redis 服务器 
 policy-map:
 api-a: #serviceId为api-a的限流规则 
 order: -1 #规则排序，但一个请求有多条规则匹配时，排序值最小的规则生效
 baseExp: Headers['userid'] #基于header里key为“userid”的值统计,目前仅支持Headers['xxx'],Cookies['xxx']
 pathRegExp: /s1/.* #请求URI的匹配正则表达式 
 timeUnit: MINUTES #时间单位，支持SECONDS,MINUTES,HOURS,DAYS
 permits: 2 #单位时间内可访问的次数
 api-a1:
 order: 0
 baseExp: Cookies['userid']
 pathRegExp: /s1.*
 timeUnit: MINUTES
 permits: 3
 api-b:
 order: 2
 baseExp: Headers['userid']
 pathRegExp: /s2/.*
 timeUnit: MINUTES
 permits: 5
eureka:
 client:
 serviceUrl:
 defaultZone: http://localhost:8761/eureka/
 instance:
 prefer-ip-address: true

参考项目zuulredislimiter/service1创建一个简单微服务

package com.tay.service1;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;

@RestController
@RequestMapping("/demo1")
public class Controller1 {

 @GetMapping("/test11")
 public String test1() {
 return "test11!";
 }
 @GetMapping("/test12")
 public String test2() {
 return "test12!";
 }
}

server:
 port: 8001
spring:
 application:
 name: service1
eureka:
 client:
 serviceUrl:
 defaultZone: http://localhost:8761/eureka/
 instance:
 prefer-ip-address: true

参考项目zuulredislimiter/service2创建一个简单微服务

package com.tay.service2;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;

@RestController
@RequestMapping("/demo2")
public class Controller2 {

 @GetMapping("/test21")
 public String test1() {
 return "test21!";
 }
 @GetMapping("/test22")
 public String test2() {
 return "test22!";
 }
}

server:
 port: 8002
spring:
 application:
 name: service2
eureka:
 client:
 serviceUrl:
 defaultZone: http://localhost:8761/eureka/
 instance:
 prefer-ip-address: true

参考项目zuulredislimiter/eurekaserver，创建一个Eureka server项目负责服务注册

package com.tay.eurekaserver;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
import org.springframework.cloud.netflix.eureka.server.EnableEurekaServer;

@EnableEurekaServer
@SpringBootApplication
public class EurekaserverApplication {
 public static void main(String[] args) {
 SpringApplication.run(EurekaserverApplication.class, args);
 }
}

server:
 port: 8761
eureka:
 client:
 register-with-eureka: false
 fetch-registry: false

Start Redis server

在本机启动Redis或者用Docker启动Redis，指令如下：

sudo docker run -d -p 6379:6379 redis

1. 启动Eureka server项目

2. 启动zuul网关项目

3. 启动微服务service1

4. 启动微服务service2

测试

你可以用一个HTTP Client工具像Postman、Restd或者curl,用GET形式访问“http://localhost:8000/s2/demo2/test21”，记得在header里面塞入userid=tom，重复请求多次，你会发现1分钟内，你最多只能成功请求5次。 通过查看配置可以得知，应该是这条限流策略生效了, 因为URI为/s2/demo2/test21跟pathRegExp: /s2/.*：匹配。

api-b:
 order: 2
 baseExp: Headers['userid']
 pathRegExp: /s2/.*
 timeUnit: MINUTES
 permits: 5

高级指南

所有的配置项

spring:
 redis-limiter: 
 redis-host: 127.0.0.1 # redis server IP default：127.0.0.1
 redis-port: 6379 # redis service port default：6379 
 redis-password: test # redis password default：null 
 redis-connection-timeout: 2000 # redis connection timeout default：2000
 redis-pool-max-idle: 50 # redis pool max idle default: 50
 redis-pool-min-idle: 10 # redis pool mim idle default：10 
 redis-pool-max-wait-millis： -1 # max wait time for get connection default：-1 
 redis-pool-max-total: 200 # max total connection default：200
 redis-key-prefix: #RL # key prefix for visit footprint default: #RL
 check-action-timeout: 100 # check action execution timeout default: 100
 channel： #RLConfigChannel # conf change event pub/sub channel default： #RLConfigChannel
 policy-map: # rate limiting policies 
 api-a: # unique service id
 order: -1 # order
 baseExp: Headers['userid'] # value to base on, Spel expression without "#", supports Headers['xxx'] or Cookies['xxx'] 
 pathRegExp: /s1/.* # URI path pattern, a Regular expression 
 timeUnit: MINUTES # timeUnit supports SECONDS, MINUTES, HOURS,DAYS
 permits: 2 # Number of visits allowed per a timeUnit
 api-a1: 
 order: 0
 baseExp: Headers['userid']
 pathRegExp: /s1.*
 timeUnit: MINUTES
 permits: 3
 ... 

动态配置

通过内置的Restful接口，我们可以动态的变更限流规则，当然这些API应该受限访问的，不过API安全性是另外一个故事，这里不展开.

@RequestMapping("/zuullimiterpolicy")
public class LimitingPolicyResource {
 ...
 @PostMapping
 public void add(@RequestBody LimitingPolicy limitingPolicy, HttpServletResponse response) throws IOException{
 ...
 }
 @PutMapping
 public void update(@RequestBody LimitingPolicy limitingPolicy, HttpServletResponse response) throws IOException {
 ...
 }
 @GetMapping("/{serviceId}")
 public LimitingPolicy get(@PathVariable("serviceId") String serviceId) {
 ...
 }
 @DeleteMapping("/{serviceId}")
 public void delete(@PathVariable("serviceId") String serviceId) {
 ...
 }

目前，限流规则支持增删查改(add, update, query, delete).

比如，在示例项目里面

GET http://localhost:8000/zuullimiterpolicy/api-a，访问值如下：

{
 "serviceId": "api-a",
 "order": -1,
 "baseExp": "Headers['userid']",
 "pathRegExp": "/s1/.*",
 "timeUnit": "MINUTES",
 "permits": 2,
 "delete": false
}

更新限流规则，指定Content-Type为“application/json”, 请求 PUT http://localhost:8000/zuullimiterpolicy, 请求request body为:

{
 "serviceId": "api-a",
 "order": -1,
 "baseExp": "Headers['userid']",
 "pathRegExp": "/s1/.*",
 "timeUnit": "MINUTES",
 "permits": 10,
 "delete": false
}

删除限流规则，规则的serviceId为"api-b"， 执行 DELETE http://localhost:8000/zuullimiterpolicy/api-b。

新增限流规则，指定Content-Type为“application/json”, 请求 POST http://localhost:8000/zuullimiterpolicy, 请求request body为:

{
 "serviceId": "api-d",
 "order": -1,
 "baseExp": "Headers['userid']",
 "pathRegExp": "/s3/.*",
 "timeUnit": "MINUTES",
 "permits": 10,
 "delete": false
}

注意，新增限流规则时，新增规则的serviceId和pathRegExp不能跟已经存在的规则相同，不然新增失败。

git地址：https://gitee.com/aiyuntang/zuul-redislimiter-spring-boot