-
springCloud的负载均衡
一.什么是负载均衡
负载均衡(Load-balance LB),指的是将用户的请求平摊分配到各个服务器上,从而达到系统的高可用。常见的负载均衡软件有Nginx、lvs等。
二.负载均衡的简单分类
1)集中式LB:集中式负载均衡指的是,在服务消费者(client)和服务提供者(provider)之间提供负载均衡设施,通过该设施把消费者(client)的请求通过某种策略转发给服务提供者(provider),常见的集中式负载均衡是Nginx;
2)进程式LB:将负载均衡的逻辑集成到消费者(client)身上,即消费者从服务注册中心获取服务列表,获知有哪些地址可用,再从这些地址里选出合适的服务器,springCloud的Ribbon就是一个进程式的负载均衡工具。
三.为什么需要做负载均衡
1) 不做负载均衡,可能导致某台机子负荷太重而挂掉;
2)导致资源浪费,比如某些机子收到太多的请求,肯定会导致某些机子收到很少请求甚至收不到请求,这样会浪费系统资源。
四.springCloud如何开启负载均衡
1)在消费者子工程的pom.xml文件的加入相关依赖(https://mvnrepository.com/artifact/org.springframework.cloud/spring-cloud-starter-ribbon/1.4.7.RELEASE);
<!-- https://mvnrepository.com/artifact/org.springframework.cloud/spring-cloud-starter-ribbon --> <dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-starter-ribbon</artifactId> <version>1.4.7.RELEASE</version> </dependency>
消费者需要获取服务注册中心的注册列表信息,把Eureka的依赖包也放进pom.xml
<dependency> <groupId>org.springframework.cloud</groupId> <artifactId>spring-cloud-starter-eureka-server</artifactId> <version>1.4.7.RELEASE</version> </dependency>
2)在application.yml里配置服务注册中心的信息
在该消费者(client)的application.yml里配置Eureka的信息,至于如何启动一个springCloud项目,请看这篇博客https://www.cnblogs.com/fengrongriup/p/14464208.html
#配置Eureka
eureka:
client:
#是否注册自己到服务注册中心,消费者不用提供服务
register-with-eureka: false
service-url:
#访问的url
defaultZone: http://localhost:8002/eureka/
3)在消费者启动类上面加上注解@EnableEurekaClient
@EnableEurekaClient
4)在配置文件的Bean上加上
@Bean @LoadBalanced public RestTemplate getRestTemplate(){ return new RestTemplate(); }
五.IRule
什么是IRule
IRule接口代表负载均衡的策略,它的不同的实现类代表不同的策略,它的四种实现类和它的关系如下()
说明一下(idea找Irule的方法:ctrl+n 填入IRule进行查找)
1.RandomRule:表示随机策略,它将从服务清单中随机选择一个服务;
public class RandomRule extends AbstractLoadBalancerRule { public RandomRule() { } @SuppressWarnings({"RCN_REDUNDANT_NULLCHECK_OF_NULL_VALUE"}) //传入一个负载均衡器 public Server choose(ILoadBalancer lb, Object key) { if (lb == null) { return null; } else { Server server = null; while(server == null) { if (Thread.interrupted()) { return null; } //通过负载均衡器获取对应的服务列表 List<Server> upList = lb.getReachableServers(); //通过负载均衡器获取全部服务列表 List<Server> allList = lb.getAllServers(); int serverCount = allList.size(); if (serverCount == 0) { return null; } //获取一个随机数 int index = this.chooseRandomInt(serverCount); //通过这个随机数从列表里获取服务 server = (Server)upList.get(index); if (server == null) { //当前线程转为就绪状态,让出cpu Thread.yield(); } else { if (server.isAlive()) { return server; } server = null; Thread.yield(); } } return server; } }
小结:通过获取到的所有服务的数量,以这个数量为标准获取一个(0,服务数量)的数作为获取服务实例的下标,从而获取到服务实例
2.ClientConfigEnabledRoundRobinRule:ClientConfigEnabledRoundRobinRule并没有实现什么特殊的处理逻辑,但是他的子类可以实现一些高级策略, 当一些本身的策略无法实现某些需求的时候,它也可以做为父类帮助实现某些策略,一般情况下我们都不会使用它;
public class ClientConfigEnabledRoundRobinRule extends AbstractLoadBalancerRule { //使用“4”中的RoundRobinRule策略 RoundRobinRule roundRobinRule = new RoundRobinRule(); public ClientConfigEnabledRoundRobinRule() { } public void initWithNiwsConfig(IClientConfig clientConfig) { this.roundRobinRule = new RoundRobinRule(); } public void setLoadBalancer(ILoadBalancer lb) { super.setLoadBalancer(lb); this.roundRobinRule.setLoadBalancer(lb); } public Server choose(Object key) { if (this.roundRobinRule != null) { return this.roundRobinRule.choose(key); } else { throw new IllegalArgumentException("This class has not been initialized with the RoundRobinRule class"); } } }
小结:用来作为父类,子类通过实现它来实现一些高级负载均衡策略
1)ClientConfigEnabledRoundRobinRule的子类BestAvailableRule:从该策略的名字就可以知道,bestAvailable的意思是最好获取的,该策略的作用是获取到最空闲的服务实例;
public class BestAvailableRule extends ClientConfigEnabledRoundRobinRule { //注入负载均衡器,它可以选择服务实例 private LoadBalancerStats loadBalancerStats; public BestAvailableRule() { } public Server choose(Object key) { //假如负载均衡器实例为空,采用它父类的负载均衡机制,也就是轮询机制,因为它的父类采用的就是轮询机制 if (this.loadBalancerStats == null) { return super.choose(key); } else { //获取所有服务实例并放入列表里 List<Server> serverList = this.getLoadBalancer().getAllServers(); //并发量 int minimalConcurrentConnections = 2147483647; long currentTime = System.currentTimeMillis(); Server chosen = null; Iterator var7 = serverList.iterator(); //遍历服务列表 while(var7.hasNext()) { Server server = (Server)var7.next(); ServerStats serverStats = this.loadBalancerStats.getSingleServerStat(server); //淘汰掉已经负载的服务实例 if (!serverStats.isCircuitBreakerTripped(currentTime)) { //获得当前服务的请求量(并发量) int concurrentConnections = serverStats.getActiveRequestsCount(currentTime); //找出并发了最小的服务 if (concurrentConnections < minimalConcurrentConnections) { minimalConcurrentConnections = concurrentConnections; chosen = server; } } } if (chosen == null) { return super.choose(key); } else { return chosen; } } } public void setLoadBalancer(ILoadBalancer lb) { super.setLoadBalancer(lb); if (lb instanceof AbstractLoadBalancer) { this.loadBalancerStats = ((AbstractLoadBalancer)lb).getLoadBalancerStats(); } } }
小结:ClientConfigEnabledRoundRobinRule子类之一,获取到并发了最少的服务
2)ClientConfigEnabledRoundRobinRule的另一个子类是PredicateBasedRule:通过源码可以看出它是一个抽象类,它的抽象方法getPredicate()返回一个AbstractServerPredicate的实例,然后它的choose方法调用AbstractServerPredicate类的chooseRoundRobinAfterFiltering方法获取具体的Server实例并返回
public abstract class PredicateBasedRule extends ClientConfigEnabledRoundRobinRule { public PredicateBasedRule() { } //获取AbstractServerPredicate对象 public abstract AbstractServerPredicate getPredicate(); public Server choose(Object key) { //获取当前策略的负载均衡器 ILoadBalancer lb = this.getLoadBalancer(); //通过AbstractServerPredicate的子类过滤掉一部分实例(它实现了Predicate) //以轮询的方式从过滤后的服务里选择一个服务 Optional<Server> server = this.getPredicate().chooseRoundRobinAfterFiltering(lb.getAllServers(), key); return server.isPresent() ? (Server)server.get() : null; } }
再看看它的chooseRoundRobinAfterFiltering()方法是如何实现的
public Optional<Server> chooseRoundRobinAfterFiltering(List<Server> servers, Object loadBalancerKey) { List<Server> eligible = this.getEligibleServers(servers, loadBalancerKey); return eligible.size() == 0 ? Optional.absent() : Optional.of(eligible.get(this.incrementAndGetModulo(eligible.size()))); }
是这样的,先通过this.getEligibleServers(servers, loadBalancerKey)方法获取一部分实例,然后判断这部分实例是否为空,如果不为空则调用eligible.get(this.incrementAndGetModulo(eligible.size())方法从这部分实例里获取一个服务,点进this.getEligibleServers看
public List<Server> getEligibleServers(List<Server> servers, Object loadBalancerKey) { if (loadBalancerKey == null) { return ImmutableList.copyOf(Iterables.filter(servers, this.getServerOnlyPredicate())); } else { List<Server> results = Lists.newArrayList(); Iterator var4 = servers.iterator(); while(var4.hasNext()) { Server server = (Server)var4.next(); //条件满足 if (this.apply(new PredicateKey(loadBalancerKey, server))) { //添加到集合里 results.add(server); } } return results; } }
getEligibleServers方法是根据this.apply(new PredicateKey(loadBalancerKey, server))进行过滤的,如果满足,就添加到返回的集合中。符合什么条件才可以进行过滤呢?可以发现,apply是用this调用的,this指的是AbstractServerPredicate(它的类对象),但是,该类是个抽象类,该实例是不存在的,需要子类去实现,它的子类在这里暂时不是看了,以后有空再深入学习下,它的子类如下,实现哪个子类,就用什么 方式过滤。
再回到chooseRoundRobinAfterFiltering()方法,刚刚说完它通过 getEligibleServers方法过滤并获取到一部分实例,然后再通过this.incrementAndGetModulo(eligible.size())方法从这部分实例里选择一个实例返回,该方法的意思是直接返回下一个整数(索引值),通过该索引值从返回的实例列表中取得Server实例。
private int incrementAndGetModulo(int modulo) { //当前下标 int current; //下一个下标 int next; do { //获得当前下标值 current = this.nextIndex.get(); next = (current + 1) % modulo; } while(!this.nextIndex.compareAndSet(current, next) || current >= modulo); return current; }
源码撸明白了,再来理一下chooseRoundRobinAfterFiltering()的思路:先通过getEligibleServers()方法获得一部分服务实例,再从这部分服务实例里拿到当前服务实例的下一个服务对象使用。
小结:通过AbstractServerPredicate的chooseRoundRobinAfterFiltering方法进行过滤,获取备选的服务实例清单,然后用线性轮询选择一个实例,是一个抽象类,过滤策略在AbstractServerPredicate的子类中具体实现
3.RetryRule:是对选定的负载均衡策略加上重试机制,即在一个配置好的时间段内(默认500ms),当选择实例不成功,则一直尝试使用subRule的方式选择一个可用的实例,在调用时间到达阀值的时候还没找到可用服务,则返回空,如果没有配置负载策略,默认轮询(即“4”中的轮询);
先贴上它的源码
public class RetryRule extends AbstractLoadBalancerRule { //从这可以看出,默认使用轮询机制 IRule subRule = new RoundRobinRule(); //阀值 long maxRetryMillis = 500L; //无参构造函数 public RetryRule() { } //使用轮询机制 public RetryRule(IRule subRule) { this.subRule = (IRule)(subRule != null ? subRule : new RoundRobinRule()); } public RetryRule(IRule subRule, long maxRetryMillis) { this.subRule = (IRule)(subRule != null ? subRule : new RoundRobinRule()); this.maxRetryMillis = maxRetryMillis > 0L ? maxRetryMillis : 500L; } public void setRule(IRule subRule) { this.subRule = (IRule)(subRule != null ? subRule : new RoundRobinRule()); } public IRule getRule() { return this.subRule; } //设置最大耗时时间(阀值),最多重试多久 public void setMaxRetryMillis(long maxRetryMillis) { if (maxRetryMillis > 0L) { this.maxRetryMillis = maxRetryMillis; } else { this.maxRetryMillis = 500L; } } //获取重试的时间 public long getMaxRetryMillis() { return this.maxRetryMillis; } //设置负载均衡器,用以获取服务 public void setLoadBalancer(ILoadBalancer lb) { super.setLoadBalancer(lb); this.subRule.setLoadBalancer(lb); } //通过负载均衡器选择服务 public Server choose(ILoadBalancer lb, Object key) { long requestTime = System.currentTimeMillis(); //当前时间+阀值 = 截止时间 long deadline = requestTime + this.maxRetryMillis; Server answer = null; answer = this.subRule.choose(key); //获取到服务直接返回 if ((answer == null || !answer.isAlive()) && System.currentTimeMillis() < deadline) { InterruptTask task = new InterruptTask(deadline - System.currentTimeMillis()); //获取不到服务的情况下反复获取 while(!Thread.interrupted()) { answer = this.subRule.choose(key); if (answer != null && answer.isAlive() || System.currentTimeMillis() >= deadline) { break; } Thread.yield(); } task.cancel(); } return answer != null && answer.isAlive() ? answer : null; } public Server choose(Object key) { return this.choose(this.getLoadBalancer(), key); } public void initWithNiwsConfig(IClientConfig clientConfig) { } }
小结:采用RoundRobinRule的选择机制,进行反复尝试,当花费时间超过设置的阈值maxRetryMills时,就返回null
4.RoundRobinRule:轮询策略,它会从服务清单中按照轮询的方式依次选择每个服务实例,它的工作原理是:直接获取下一个可用实例,如果超过十次没有获取到可用的服务实例,则返回空且报出异常信息;
public class RoundRobinRule extends AbstractLoadBalancerRule { private AtomicInteger nextServerCyclicCounter; private static final boolean AVAILABLE_ONLY_SERVERS = true; private static final boolean ALL_SERVERS = false; private static Logger log = LoggerFactory.getLogger(RoundRobinRule.class); public RoundRobinRule() { this.nextServerCyclicCounter = new AtomicInteger(0); } public RoundRobinRule(ILoadBalancer lb) { this(); this.setLoadBalancer(lb); } public Server choose(ILoadBalancer lb, Object key) { if (lb == null) { log.warn("no load balancer"); return null; } else { Server server = null; int count = 0; while(true) { //选择十次,十次都没选到可用服务就返回空 if (server == null && count++ < 10) { List<Server> reachableServers = lb.getReachableServers(); List<Server> allServers = lb.getAllServers(); int upCount = reachableServers.size(); int serverCount = allServers.size(); if (upCount != 0 && serverCount != 0) { int nextServerIndex = this.incrementAndGetModulo(serverCount); server = (Server)allServers.get(nextServerIndex); if (server == null) { Thread.yield(); } else { if (server.isAlive() && server.isReadyToServe()) { return server; } server = null; } continue; } log.warn("No up servers available from load balancer: " + lb); return null; } if (count >= 10) { log.warn("No available alive servers after 10 tries from load balancer: " + lb); } return server; } } } //递增的形式实现轮询 private int incrementAndGetModulo(int modulo) { int current; int next; do { current = this.nextServerCyclicCounter.get(); next = (current + 1) % modulo; } while(!this.nextServerCyclicCounter.compareAndSet(current, next)); return next; } public Server choose(Object key) { return this.choose(this.getLoadBalancer(), key); } public void initWithNiwsConfig(IClientConfig clientConfig) { } }
小结:采用线性轮询机制循环依次选择每个服务实例,直到选择到一个不为空的服务实例或循环次数达到10次
它有个子类WeightedResponseTimeRule,WeightedResponseTimeRule是对RoundRobinRule的优化。WeightedResponseTimeRule在其父类的基础上,增加了定时任务这个功能,通过启动一个定时任务来计算每个服务的权重,然后遍历服务列表选择服务实例,从而达到更加优秀的分配效果。我们这里把这个类分为三部分:定时任务,计算权值,选择服务
1)定时任务
//定时任务 void initialize(ILoadBalancer lb) { if (this.serverWeightTimer != null) { this.serverWeightTimer.cancel(); } this.serverWeightTimer = new Timer("NFLoadBalancer-serverWeightTimer-" + this.name, true); //开启一个任务,每30秒执行一次 this.serverWeightTimer.schedule(new WeightedResponseTimeRule.DynamicServerWeightTask(), 0L, (long)this.serverWeightTaskTimerInterval); WeightedResponseTimeRule.ServerWeight sw = new WeightedResponseTimeRule.ServerWeight(); sw.maintainWeights(); Runtime.getRuntime().addShutdownHook(new Thread(new Runnable() { public void run() { WeightedResponseTimeRule.logger.info("Stopping NFLoadBalancer-serverWeightTimer-" + WeightedResponseTimeRule.this.name); WeightedResponseTimeRule.this.serverWeightTimer.cancel(); } })); }
DynamicServerWeightTask()任务如下:
class DynamicServerWeightTask extends TimerTask { DynamicServerWeightTask() { } public void run() { WeightedResponseTimeRule.ServerWeight serverWeight = WeightedResponseTimeRule.this.new ServerWeight(); try { //计算权重 serverWeight.maintainWeights(); } catch (Exception var3) { WeightedResponseTimeRule.logger.error("Error running DynamicServerWeightTask for {}", WeightedResponseTimeRule.this.name, var3); } } }
小结:调用initialize方法开启定时任务,再在任务里计算服务的权重
2)计算权重:第一步,先算出所有实例的响应时间;第二步,再根据所有实例响应时间,算出每个实例的权重
//用来存储权重 private volatile List<Double> accumulatedWeights = new ArrayList(); //内部类 class ServerWeight { ServerWeight() { } //该方法用于计算权重 public void maintainWeights() { //获取负载均衡器 ILoadBalancer lb = WeightedResponseTimeRule.this.getLoadBalancer(); if (lb != null) { if (WeightedResponseTimeRule.this.serverWeightAssignmentInProgress.compareAndSet(false, true)) { try { WeightedResponseTimeRule.logger.info("Weight adjusting job started"); AbstractLoadBalancer nlb = (AbstractLoadBalancer)lb; //获得每个服务实例的信息 LoadBalancerStats stats = nlb.getLoadBalancerStats(); if (stats != null) { //实例的响应时间 double totalResponseTime = 0.0D; ServerStats ss; //累加所有实例的响应时间 for(Iterator var6 = nlb.getAllServers().iterator(); var6.hasNext(); totalResponseTime += ss.getResponseTimeAvg()) { Server server = (Server)var6.next(); ss = stats.getSingleServerStat(server); } Double weightSoFar = 0.0D; List<Double> finalWeights = new ArrayList(); Iterator var20 = nlb.getAllServers().iterator(); //计算负载均衡器所有服务的权重,公式是weightSoFar = weightSoFar + weight-实例平均响应时间 while(var20.hasNext()) { Server serverx = (Server)var20.next(); ServerStats ssx = stats.getSingleServerStat(serverx); double weight = totalResponseTime - ssx.getResponseTimeAvg(); weightSoFar = weightSoFar + weight; finalWeights.add(weightSoFar); } WeightedResponseTimeRule.this.setWeights(finalWeights); return; } } catch (Exception var16) { WeightedResponseTimeRule.logger.error("Error calculating server weights", var16); return; } finally { WeightedResponseTimeRule.this.serverWeightAssignmentInProgress.set(false); } } } } }
3)选择服务
@SuppressWarnings({"RCN_REDUNDANT_NULLCHECK_OF_NULL_VALUE"}) public Server choose(ILoadBalancer lb, Object key) { if (lb == null) { return null; } else { Server server = null; while(server == null) { List<Double> currentWeights = this.accumulatedWeights; if (Thread.interrupted()) { return null; } List<Server> allList = lb.getAllServers(); int serverCount = allList.size(); if (serverCount == 0) { return null; } int serverIndex = 0; double maxTotalWeight = currentWeights.size() == 0 ? 0.0D : (Double)currentWeights.get(currentWeights.size() - 1); if (maxTotalWeight >= 0.001D && serverCount == currentWeights.size()) { //生产0到最大权重值的随机数 double randomWeight = this.random.nextDouble() * maxTotalWeight; int n = 0; //循环权重区间 for(Iterator var13 = currentWeights.iterator(); var13.hasNext(); ++n) { //获取到循环的数 Double d = (Double)var13.next(); //假如随机数在这个区间内,就拿该索引d服务列表获取对应的实例 if (d >= randomWeight) { serverIndex = n; break; } } server = (Server)allList.get(serverIndex); } else { server = super.choose(this.getLoadBalancer(), key); if (server == null) { return server; } } if (server == null) { Thread.yield(); } else { if (server.isAlive()) { return server; } server = null; } } return server; } }
小结:首先生成了一个[0,最大权重值) 区间内的随机数,然后遍历权重列表,假如当前随机数在这个区间内,就通过该下标获得对应的服务。
原文:https://www.cnblogs.com/fengrongriup/p/14505755.html