SpringCloud之Hystrix的详细使用

目录

1.概念

服务降级：服务器繁忙，请稍后再试，不让客户端等待，并立即返回一个友好的提示（一般发生在 程序异常，超时，服务熔断触发服务降级，线程池、信号量 打满也会导致服务降级）
服务熔断 ： 达到最大服务访问后，直接拒绝访问，然后调用服务降级的方法并返回友好提示（如保险丝一样）
服务限流 ： 秒杀等高并发操作，严禁一窝蜂的过来拥挤，排队进入，一秒钟n个，有序进行

***一.服务降级***

2.不使用hystrix的项目

大致的rvice和controller层如下

***************control宁波水上乐园ler*****************package com.sky.springcloud.controller;import com.sky.springcloud.rvice.paymentrvice;import lombok.extern.slf4j.slf4j;import org.springframework.beans.factory.annotation.value;import org.springframework.web.bind.annotation.getmapping;import org.springframework.web.bind.annotation.pathvariable;import org.springframework.web.bind.annotation.restcontroller;import javax.annotation.resource;@restcontroller@slf4jpublic class paymentcontroller {    @resource    private paymentrvice paymentrvice;    @value("${rver.port}")    private string rverport;    @getmapping("/payment/hystrix/ok/{id}")    public string paymentinfo_ok(@pathvariable("id") integer id){        string result = paymentrvice.paymentinfo_ok(id);        log.info("*****"+ result);        return result;    }    @getmapping("/payment/hystrix/timeout/{id}")    public string paymentinfo_timeout(@pathvariable("id") integer id){        string result = paymentrvice.payment_info_timeout(id);}*******************rvice******************** package com.sky.springcloud.rvice;import org.springframework.stereotype.rvice;import java.util.concurrent.timeunit;@rvicepublic class paymentrviceimpl implements paymentrvice{    @override    public string paymentinfo_ok(integer id) {        return "线程池：" + thread.currentthread().getname() + "paymentinfo_ok. id:" + id + "\t" + "~~~~~";    public string payment_info_timeout(integer id) {       try {            timeunit.conds.sleep(5);        } catch (interruptedexception e) {            e.printstacktrace();        }        return "线程池：" + thread.currentthread().getname() + "paymentinfo_timeout. id:" + id + "\t" + "~~~~~";

在这种情况时,当通过浏览器访问 timeout这个方法,会三秒后返回结果,而当访问 ok 这个方法时,会直接返回结果(但是这是在访问量很少的时候,一旦访问量过多,访问ok时也会出现延迟,这里可以使用jmeter模拟两万个访问量,如下)

使用 jmeter 模拟后,再去访问ok,会明显出现加载的效果

3. 使用hystrix

在主启动类添加注解
@enablehystrix

package com.sky.springcloud;import org.spri杭州艺术学校ngframework.boot.springapplication;import org.springframework.boot.autoconfigure.springbootapplication;import org.springframework.cloud.netflix.eureka.enableeurekaclient;import org.springframework.cloud.netflix.hystrix.enablehystrix;@springbootapplication@enableeurekaclient //启用eureka@enablehystrix     //启用hystrixpublic class payment8001 {    public static void main(string[] args) {        springapplication.run(payment8001.class,args);    }}

在原有的基础上,在rvice层加上如下注解

 @hystrixcommand(fallbackmethod = "payment_info_timeouthandler",//当服务降级时,调用payment_info_timeouthandler方法                  commandproperties = {            @hystrixproperty(name = "execution.isolation.thread.timeoutinmilliconds",            value = "3000")//设置时间为3秒,超过三秒就为时间超限

*****************改后的rvice*******************package com.sky.springcloud.rvice;import com.netflix.hystrix.contrib.javanica.annotation.hystrixcommand;import com.netflix.hystrix.contrib.javanica.annotation.hystrixproperty;import org.springframework.stereotype.rvice;import java.util.concurrent.timeunit;@rvicepublic class paymentrviceimpl implements paymentrvice{    @override    public string paymentinfo_ok(integer id) {        return "线程池：" + thread.currentthread().getname() + "paymentinfo_ok. id:" + id + "\t" + "~~~~~";    }    @hystrixcommand(fallbackmethod = "payment_info_timeouthandler",commandproperties = {            @hystrixproperty(name = "execution.isolation.thread.timeoutinmilliconds",value = "3000")    })    public string payment_info_timeout(integer id) {      try {            timeunit.conds.sleep(5);        } catch (interruptedexception e) {            e.printstacktrace();        }      // int a = 10 / 0;        return "线程池：" + thread.currentthread().getname() + "paymentinfo_timeout. id:" + id + "\t" + "~~~~~";    }    public string payment_info_timeouthandler(integer id){        return "线程超时或异常 " + thread.currentthread().getname();    }}

如上rvice所示,如果再次访问timeout需要等待五秒,但是hystrix设置的超时时间为三秒,所以当三秒后没有结果就会服务降级,从而调用timeouthandler这个指定 的方法来执行(或者有程序出错等情况也会服务降级)

4. 全局的hystrix配置

@defaultproperties(defaultfallback = “gloub_test”)

package com.sky.springcloud.rvice;import com.netflix.hystrix.contrib.javanica.annotation.defaultproperties;import com.netflix.hystrix.contrib.javanica.annotation.hystrixcommand;import com.netflix.hystrix.contrib.javanica.annotation.hystrixproperty;import org.springframework.stereotype.rvice;@rvice@defaultproperties(defaultfallback = "gloub_test")public class paymentrviceimpl implements paymentrvice{    @override    @hystrixcommand    public string paymentinfo_ok(integer id) {        int a = 10 / 0;        return "线程池：" + thread.currentthread().getname() + "paymentinfo_ok. id:" + id + "\t" + "~~~~~";    }    @hystrixcommand(fallbackmethod = "payment_info_timeouthandler",commandproperties = {            @hystrixproperty(name = "execution.isolation.thread.timeoutinmilliconds",value = "3000")    })    public string payment_info_timeout(integer id) {   /*     try {            timeunit.conds.sleep(5);        } catch (interruptedexception e) {            e.printstacktrace();        }*/    和颜    return "线程池：" + thread.currentthread().getname() + "paymentinfo_timeout. id:" + id + "\t" + "~~~~~";    public string payment_info_timeouthandler(integer id){        return "线程超时或异常 " + thread.currentthread().getname();    public string gloub_test(){        return "走了全局的";}

如上rvice所示,加了@hystrixcommand的方金水银水法里,
如果没指定fallbackmethod 就会默认去找全局的defaultfallback
这里 当paymentinfo_ok 方法出错时,会调用gloub_test方法
当payment_info_timeout出错时,会调用payment_info_timeouthandler方法

***二.服务熔断***

1.熔断机制概述

熔断机制是应对雪崩效应的一种微服务链路保护机制,当扇出链路的某个微服务出错不可用或者响应时间太长,会进行服务的降级,进而熔断该节点微服务的调用,快速返回错误的相应信息.当检测到该节点微服务调用相应正常后,恢复调用链路.

2.项目中使用

在上面的基础上,改写rvice和controller(添加以下代码)

//是否开启服务熔断(断路器) @hystrixproperty(name = "circuitbreaker.enabled",value = "true"), //服务请求的次数  @hystrixproperty(name = "circuitbreaker.requestvolumethreshold",value = "10"), //时间的窗口期 @hystrixproperty(name = "circuitbreaker.sleepwindowinmilliconds", value = "10000"), //当失败率达到多少后发生熔断 @hystrixproperty(name = "circuitbreaker.errorthresholdpercentage",value = "60"), ******************rvice*****************    @hystrixcommand(fallbackmethod = "paymentcircuitbreaker_fallback",                    commandproperties = {                    @hystrixproperty(name = "circuitbreaker.enabled",value = "true"),                    @hystrixproperty(name = "circuitbreaker.requestvolumethreshold",value = "10"),                    @hystrixproperty(name = "circuitbreaker.sleepwindowinmilliconds", value = "10000"),                    @hystrixproperty(name = "circuitbreaker.errorthresholdpercentage",value = "60"),                    })    public string paymentcircuitbreaker(@pathvariable("id") integer id){        if(id<0){            throw new runtimeexception("********** id不能为负");        }        string rialnumber = idutil.simpleuuid();        return thread.currentthread().getname() + "\t" + rialnumber;    }    public string paymentcircuitbreaker_fallback(@pathvariable("id") integer id){        return  "id 不能为负:" + id;************controller********************    @getmapping("/payment/circuit/{id}")        string result = paymentrvice.paymentcircuitbreaker(id);        log.info(result + "***********");        return result;

如上所示,当访问paymentcircuitbreaker时,如果id小于零则会有一个运行错误,这时候就会通过服务降级来调用paymentcircuitbreaker_fallback方法,当错误的次数达到60%的时候(自己设的),就会出现服务熔断,这个时候再访问id大于零的情况,也会发生服务降级,因为开起了服务熔断,需要慢慢的恢复正常.

到此这篇关于springcloud之hystrix的详细使用的文章就介绍到这了,更多相关springcloud hystrix使用内容请搜索www.887551.com以前的文章或继续浏览下面的相关文章希望大家以后多多支持www.887551.com！