hystrix 源码分析以及属性的配置

一.feign与hystix结合

1.1测试环境搭建

架构如图：

非常简单，就是Order服务通过feign调用product服务的一个获取商品信息的一个接口：

package com.yang.xiao.hui.order.controller;

import org.springframework.cloud.openfeign.FeignClient;
import org.springframework.context.annotation.Primary;
import org.springframework.web.bind.annotation.PathVariable;
import org.springframework.web.bind.annotation.RequestMapping;

@FeignClient(name ="product",fallback =ProductHystrix.class)
@Primary
public interface ProductService {
    @RequestMapping("/info/{id}")
    Product getProductInfo(@PathVariable("id") Integer id);
}


@Component
public class ProductHystrix implements ProductService{
    @Override
    public Product getProductInfo(Integer id) {
        System.out.println("被熔断；了");
        Product product = new Product();
        product.setName("熔断了。。。。。");

        return product;
    }
}

我们在order服务引入feign的依赖：

     <dependency>
            <groupId>org.springframework.cloud</groupId>
            <artifactId>spring-cloud-starter-openfeign</artifactId>
        </dependency>

通过下面的依赖关系图可见，已经引入了hystrix相关依赖：

根据FeignClientsConfiguration这个配置类的一个方法feignHystrixBuilder()，可以知道，要让feign跟hystrix结合，需要在application.yml配置一个feign.hystrix.enabled =true的属性

1.2 debug调试feignClient客户端的创建：

.根据上次feign的调用源码分析，可以知道，带有@FeignClient的接口都会被封装成FeignClientFactoryBean类，通过该类的getObject()方法可以获取对应的代理对象，具体源码分析，参考https://www.cnblogs.com/yangxiaohui227/p/12965340.html

断点打在FeignClientFactoryBean的getObject（）方法，然后启动Order服务：（因为我的order服务的Controller注入ProductSerivce,所以启动时，会调用FeignClientFactoryBean的getObject（）获取实例）

@Controller
public class OrderContorller {
    @Autowired
    private ProductService productService;


    @ResponseBody
    @RequestMapping("/info/{id}")
    @MyLogAnnotation
    public Product getProductInfo(@PathVariable Integer id){
        return productService.getProductInfo(id);
    }
}

上面先调用build，在调用newInstance来创建代理对象，先看build方法：

至此代理对象创建完毕，我们看到了方法处理器是：HystrixInvocationHandler

1.3 debug调试feign调用第三方接口：

在浏览器输入http://localhost:8674/info/3,调用order服务的getProductInfo方法：

debug跟踪进去：

可见，最终是进入了HystrixInvocationHandler类的invoke方法：方法具体如下：

 @Override
  public Object invoke(final Object proxy, final Method method, final Object[] args)
      throws Throwable {
    // early exit if the invoked method is from java.lang.Object
    // code is the same as ReflectiveFeign.FeignInvocationHandler
    if ("equals".equals(method.getName())) {
      try {
        Object otherHandler =
            args.length > 0 && args[0] != null ? Proxy.getInvocationHandler(args[0]) : null;
        return equals(otherHandler);
      } catch (IllegalArgumentException e) {
        return false;
      }
    } else if ("hashCode".equals(method.getName())) {
      return hashCode();
    } else if ("toString".equals(method.getName())) {
      return toString();
    }

    //这里创建一个command命令对象，里面实现了2个方法，一个是执行我们目标方法的run()方法，一个是执行目标方法失败后，调用我们的服务降级方法getFallcack()
    HystrixCommand<Object> hystrixCommand =
        new HystrixCommand<Object>(setterMethodMap.get(method)) {
          @Override
          protected Object run() throws Exception {
            try {
              return HystrixInvocationHandler.this.dispatch.get(method).invoke(args);
            } catch (Exception e) {
              throw e;
            } catch (Throwable t) {
              throw (Error) t;
            }
          }

          @Override
          protected Object getFallback() {
            if (fallbackFactory == null) {
              return super.getFallback();
            }
            try {
              Object fallback = fallbackFactory.create(getExecutionException());
              Object result = fallbackMethodMap.get(method).invoke(fallback, args);
              if (isReturnsHystrixCommand(method)) {
                return ((HystrixCommand) result).execute();
              } else if (isReturnsObservable(method)) {
                // Create a cold Observable
                return ((Observable) result).toBlocking().first();
              } else if (isReturnsSingle(method)) {
                // Create a cold Observable as a Single
                return ((Single) result).toObservable().toBlocking().first();
              } else if (isReturnsCompletable(method)) {
                ((Completable) result).await();
                return null;
              } else if (isReturnsCompletableFuture(method)) {
                return ((Future) result).get();
              } else {
                return result;
              }
            } catch (IllegalAccessException e) {
              // shouldn't happen as method is public due to being an interface
              throw new AssertionError(e);
            } catch (InvocationTargetException | ExecutionException e) {
              // Exceptions on fallback are tossed by Hystrix
              throw new AssertionError(e.getCause());
            } catch (InterruptedException e) {
              // Exceptions on fallback are tossed by Hystrix
              Thread.currentThread().interrupt();
              throw new AssertionError(e.getCause());
            }
          }
        };

    if (Util.isDefault(method)) {
      return hystrixCommand.execute();
    } else if (isReturnsHystrixCommand(method)) {
      return hystrixCommand;
    } else if (isReturnsObservable(method)) {
      // Create a cold Observable
      return hystrixCommand.toObservable();
    } else if (isReturnsSingle(method)) {
      // Create a cold Observable as a Single
      return hystrixCommand.toObservable().toSingle();
    } else if (isReturnsCompletable(method)) {
      return hystrixCommand.toObservable().toCompletable();
    } else if (isReturnsCompletableFuture(method)) {
      return new ObservableCompletableFuture<>(hystrixCommand);
    }
    return hystrixCommand.execute(); //最终调用了命令行的execute方法
  }

小结：这里创建一个command命令对象，里面实现了2个方法，一个是执行我们目标方法的run()方法，一个是执行目标方法失败后，调用我们的服务降级方法getFallcack()，直接调用了命令对象的execute方法：1.4 我们跟进命令对象的创建看看有哪些属性：

这里一大堆初始化属性，我们主要关注2个：HystrixCommandProperties.Setter 和 HystrixThreadPoolProperties.Setter，以及线程池的初始化

我们看到线程池的初始化，有个threadPoolkey,该值为服务的名称，也就是@FeignClient注解的name 值，初始化时，同时传入了一个线程池的默认配置：

那么我们先看HystrixThreadPoolProperties.Setter的属性有哪些值：

线程池属性初始化前缀是hystrix，通过上面我们可以知道，Hystrix有默认的线程参数，超过对应的线程参数就会被拒绝，那么我们就可以用该特征来做服务限流了，真对某个方法给其设置相应的最大队列属性；

直接我们查看HystrixCommandProperties的属性值：

命令创建分析完后，我们跟踪执行逻辑：hystrixCommand.execute();

这里用到了大量的RXJava相关知识，从上面代码可以知道，该方法是阻塞的：

跟进toObservable()，我们找到核心代码：

private Observable<R> applyHystrixSemantics(final AbstractCommand<R> _cmd) {
        // mark that we're starting execution on the ExecutionHook
        // if this hook throws an exception, then a fast-fail occurs with no fallback.  No state is left inconsistent
        executionHook.onStart(_cmd);

        /* determine if we're allowed to execute */
        if (circuitBreaker.allowRequest()) { //这里判断是否允许执行目标方法，如果不允许，那就执行fallbcak方法
            final TryableSemaphore executionSemaphore = getExecutionSemaphore();
            final AtomicBoolean semaphoreHasBeenReleased = new AtomicBoolean(false);
            final Action0 singleSemaphoreRelease = new Action0() {
                @Override
                public void call() {
                    if (semaphoreHasBeenReleased.compareAndSet(false, true)) {
                        executionSemaphore.release();
                    }
                }
            };

            final Action1<Throwable> markExceptionThrown = new Action1<Throwable>() {
                @Override
                public void call(Throwable t) {
                    eventNotifier.markEvent(HystrixEventType.EXCEPTION_THROWN, commandKey);
                }
            };

            if (executionSemaphore.tryAcquire()) {
                try {
                    /* used to track userThreadExecutionTime */
                    executionResult = executionResult.setInvocationStartTime(System.currentTimeMillis());
                    return executeCommandAndObserve(_cmd)//这里是执行目标方法
                            .doOnError(markExceptionThrown)
                            .doOnTerminate(singleSemaphoreRelease)
                            .doOnUnsubscribe(singleSemaphoreRelease);
                } catch (RuntimeException e) {
                    return Observable.error(e);
                }
            } else {
                return handleSemaphoreRejectionViaFallback();//处理fallback方法
            }
        } else {
            return handleShortCircuitViaFallback();//处理fallback方法
        }
    }

重点三个方法：circuitBreaker.allowRequest() /executeCommandAndObserve(_cmd) / handleSemaphoreRejectionViaFallback()

先看熔断器核心业务方法：circuitBreaker.allowRequest()

我们看看isOpen()方法的逻辑：

       @Override
        public boolean isOpen() {
            if (circuitOpen.get()) { //如果已经是打开的，那就返回true
                // if we're open we immediately return true and don't bother attempting to 'close' ourself as that is left to allowSingleTest and a subsequent successful test to close
                return true;
            }

            // we're closed, so let's see if errors have made us so we should trip the circuit open
            HealthCounts health = metrics.getHealthCounts(); //获取统计对象，每个command对象的key值就会有一个统计对象，也就是每个请求方法对应一个对象

            // check if we are past the statisticalWindowVolumeThreshold
            if (health.getTotalRequests() < properties.circuitBreakerRequestVolumeThreshold().get()) { //在一个时间窗口内，默认10s,请求总次数要达到配置的最小请求数（默认20次）
                // we are not past the minimum volume threshold for the statisticalWindow so we'll return false immediately and not calculate anything
                return false;
            }

            if (health.getErrorPercentage() < properties.circuitBreakerErrorThresholdPercentage().get()) { /在一个时间窗口内，默认10s,请求失败率要达到最小配置百分比（默认50%）
                return false;
            } else {
                // our failure rate is too high, trip the circuit //执行到这里说明失败率过高
                if (circuitOpen.compareAndSet(false, true)) {
                    // if the previousValue was false then we want to set the currentTime //即使之前是关闭状态也要将其改为打开状态
                    circuitOpenedOrLastTestedTime.set(System.currentTimeMillis());//设置最近打开时间，用于半打开状态的设置
                    return true;
                } else {
                    // How could previousValue be true? If another thread was going through this code at the same time a race-condition could have
                    // caused another thread to set it to true already even though we were in the process of doing the same
                    // In this case, we know the circuit is open, so let the other thread set the currentTime and report back that the circuit is open
                    return true;
                }
            }
        }

    }

小结：isOpen()在默认情况下，如果10s内请求次数达到20次，失败率达到百分之50，则熔断器将会打开

接着我们分析半打开方法：allowSingleTest()；

public boolean allowSingleTest() {
            long timeCircuitOpenedOrWasLastTested = circuitOpenedOrLastTestedTime.get(); //上次最近打开熔断器的时间
            // 1) if the circuit is open
            // 2) and it's been longer than 'sleepWindow' since we opened the circuit

           //熔断器如果是打开状态，并且最近打开时间距离当前时间已经超过了设置的时间（默认5s）

           if (circuitOpen.get() && System.currentTimeMillis() > timeCircuitOpenedOrWasLastTested + properties.circuitBreakerSleepWindowInMilliseconds().get()) { 
                // We push the 'circuitOpenedTime' ahead by 'sleepWindow' since we have allowed one request to try.
                // If it succeeds the circuit will be closed, otherwise another singleTest will be allowed at the end of the 'sleepWindow'.
                if (circuitOpenedOrLastTestedTime.compareAndSet(timeCircuitOpenedOrWasLastTested, System.currentTimeMillis())) { //更新最近打开时间，方便下一个5s进行判断
                    // if this returns true that means we set the time so we'll return true to allow the singleTest
                    // if it returned false it means another thread raced us and allowed the singleTest before we did
                    return true;
                }
            }
            return false;
        }

一张图总结：

至此，熔断器核心逻辑分析完毕了，我们回到之前目标方法的逻辑，继续跟踪executeCommandAndObserve(_cmd)；

前面我们分析过，创建Command对象时，实现了run方法，和getFallback()方法，上图可见，run方法被调用了，而run方法里面有我们的目标方法

接着我们分析执行目标方法失败后，是如何调用getFallback方法的： handleSemaphoreRejectionViaFallback()：

由此可见，这里最终调用了Command对象的getFallback（）方法；

执行失败的原因有：执行异常，超时，熔断器打开，线程数超过最大限制，请求异常

1.4 如何修改默认的参数配置：如，我想更改线程池的配置，或者滑动窗口时间，@FeignClient注解里面并没有更改hystrix的相关配置，那么我们回顾下ProductService接口的创建过程：

我们创建命令对象时：

由此可知，我们创建的HystrixCommand时，要用到一个属性配置对象，该对象其实就是创建代理对象时，由HystrixFeign提供，但该属性是直接new出来的，所以我们唯有替换掉HystrixFeign

我们发现，容器中没有HystrixFeign.Builder对象时就才会创建Feign.Builder，那么我们就给他注入一股Feign.Builder，在里面给它设置setterFactory属性：

package com.yang.xiao.hui.order.controller;

import com.netflix.hystrix.HystrixCommand;
import com.netflix.hystrix.HystrixCommandGroupKey;
import com.netflix.hystrix.HystrixCommandKey;
import com.netflix.hystrix.HystrixThreadPoolProperties;
import feign.Feign;
import feign.Target;
import feign.hystrix.HystrixFeign;
import feign.hystrix.SetterFactory;
import org.springframework.boot.autoconfigure.condition.ConditionalOnClass;
import org.springframework.boot.autoconfigure.condition.ConditionalOnProperty;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Scope;

import java.lang.reflect.Method;


@Configuration(proxyBeanMethods = false)
@ConditionalOnClass({ HystrixCommand.class, HystrixFeign.class })
public class FeignConfig {

    @Bean
    @Scope("prototype")
    @ConditionalOnProperty(name = "feign.hystrix.enabled")
    public Feign.Builder feignHystrixBuilder() {
        HystrixFeign.Builder builder = HystrixFeign.builder();
        SetterFactory setterFactory= new SetterFactory(){

            @Override
            public HystrixCommand.Setter create(Target<?> target, Method method) {
                String groupKey = target.name();
                String commandKey = Feign.configKey(target.type(), method);
                return HystrixCommand.Setter
                        .withGroupKey(HystrixCommandGroupKey.Factory.asKey(groupKey))
                        .andCommandKey(HystrixCommandKey.Factory.asKey(commandKey))
                        .andThreadPoolPropertiesDefaults(HystrixThreadPoolProperties.Setter().withCoreSize(100).withMaxQueueSize(200));

            }
        };
        builder.setterFactory(setterFactory);

        return builder;
    }
}

//需要更改任何属性就在HystrixCommand.Setter对象加，我们这里更改了线程池的核心线程数为100，最大队列数为200，启动order服务，重新调用方法：

已经变更了；

思考：上面的配置是对所有的服务都起效的，如果我们需要对不同的服务，如库存服务，客户服务，等有不同的配置，如何处理：

方案一，在主启动类中，将上面的配置类的扫描排除，然后在@FeignClient注解中添加该配置，原理在feign源码里说过，在创建feign对象时，如果子容器不存在某个bean,会到父，容器中找；那么我们的思路是让父容器扫描不到：

此时再次测试：

方案二：在配置类那里根据不同服务返回不同设置，甚至可以根据不同方法返回不同设置，因为Feign.Builder对象是多例的：

package com.yang.xiao.hui.order.controller;

import com.netflix.hystrix.HystrixCommand;
import com.netflix.hystrix.HystrixCommandGroupKey;
import com.netflix.hystrix.HystrixCommandKey;
import com.netflix.hystrix.HystrixThreadPoolProperties;
import feign.Feign;
import feign.Target;
import feign.hystrix.HystrixFeign;
import feign.hystrix.SetterFactory;
import org.springframework.boot.autoconfigure.condition.ConditionalOnClass;
import org.springframework.boot.autoconfigure.condition.ConditionalOnProperty;
import org.springframework.context.annotation.Bean;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Scope;

import java.lang.reflect.Method;


@Configuration(proxyBeanMethods = false)
@ConditionalOnClass({ HystrixCommand.class, HystrixFeign.class })
public class FeignConfig {

    @Bean
    @Scope("prototype")
    @ConditionalOnProperty(name = "feign.hystrix.enabled")
    public Feign.Builder feignHystrixBuilder() {
        HystrixFeign.Builder builder = HystrixFeign.builder();
        SetterFactory setterFactory= new SetterFactory(){

            @Override
            public HystrixCommand.Setter create(Target<?> target, Method method) {
                String groupKey = target.name();
                String commandKey = Feign.configKey(target.type(), method);
                HystrixThreadPoolProperties.Setter setter = HystrixThreadPoolProperties.Setter().withCoreSize(100).withMaxQueueSize(200);
                if("product".equals(groupKey)){ //如果是商品服务就用下面配置
                    setter=HystrixThreadPoolProperties.Setter().withCoreSize(50).withMaxQueueSize(50);
                }
                return HystrixCommand.Setter
                        .withGroupKey(HystrixCommandGroupKey.Factory.asKey(groupKey))
                        .andCommandKey(HystrixCommandKey.Factory.asKey(commandKey))
                        .andThreadPoolPropertiesDefaults(setter);

            }
        };
        builder.setterFactory(setterFactory);

        return builder;
    }
}

二.hystix用于controller层

前面分析的是feign与hystrix结合的源码，hystix还可以用于controller层：

2.1 添加依赖：

      <dependency>
            <groupId>org.springframework.cloud</groupId>
            <artifactId>spring-cloud-starter-netflix-hystrix</artifactId>
        </dependency>

2.2 修改controller