Sentinel的限流、熔断源码浅析

一、简单使用（源码分析基石）

1.引入 Sentinel 依赖

		<dependency>
			<groupId>com.alibaba.cloud</groupId>
			<artifactId>spring-cloud-starter-alibaba-sentinel</artifactId>
		</dependency>

2.对Controller方法打上注解

	@SentinelResource(value = "/order/create", blockHandler = "blockHandlerFunc", fallback = "fallbackFunc")
    @GetMapping("/test-sentinel-resource")
    public String testSentinelResource(@RequestParam(required = false) String a)
            throws InterruptedException {
        // 模拟执行被保护的业务逻辑耗时
        Thread.sleep(100);
        return a;
    }

二、注解解析-----切面

通过一个切面类SentinelResourceAspect，去解析我们写的@SentinelResource

@Aspect
public class SentinelResourceAspect extends AbstractSentinelAspectSupport {

	//切点，SentinelResource注解
    @Pointcut("@annotation(com.alibaba.csp.sentinel.annotation.SentinelResource)")
    public void sentinelResourceAnnotationPointcut() {
    }

	//环绕通知
    @Around("sentinelResourceAnnotationPointcut()")
    public Object invokeResourceWithSentinel(ProceedingJoinPoint pjp) throws Throwable {
        Method originMethod = resolveMethod(pjp);

        SentinelResource annotation = originMethod.getAnnotation(SentinelResource.class);
        if (annotation == null) {
            // Should not go through here.
            throw new IllegalStateException("Wrong state for SentinelResource annotation");
        }
        String resourceName = getResourceName(annotation.value(), originMethod);
        EntryType entryType = annotation.entryType();
        int resourceType = annotation.resourceType();
        Entry entry = null;
        try {
            entry = SphU.entry(resourceName, resourceType, entryType, pjp.getArgs());
            Object result = pjp.proceed();
            return result;
        } catch (BlockException ex) {
            return handleBlockException(pjp, annotation, ex);
        } catch (Throwable ex) {
            Class<? extends Throwable>[] exceptionsToIgnore = annotation.exceptionsToIgnore();
            // The ignore list will be checked first.
            if (exceptionsToIgnore.length > 0 && exceptionBelongsTo(ex, exceptionsToIgnore)) {
                throw ex;
            }
            if (exceptionBelongsTo(ex, annotation.exceptionsToTrace())) {
                traceException(ex, annotation);
                return handleFallback(pjp, annotation, ex);
            }
            throw ex;
        } finally {
            if (entry != null) {
                entry.exit(1, pjp.getArgs());
            }
        }
    }
}

三、处理方式----try/catch

通过一个大大的try/catch，包裹住我们的方法，在之前执行SphU.entry(通过在这规则校验抛出异常而阻止后续方法正常执行达到限流熔断)，执行之后根据方法执行的情况进行记录处理等：

        try {
            entry = SphU.entry(resourceName, resourceType, entryType, pjp.getArgs());
            Object result = pjp.proceed();
            return result;
        } catch (BlockException ex) {
            return handleBlockException(pjp, annotation, ex);
        } catch (Throwable ex) {
            if (exceptionBelongsTo(ex, annotation.exceptionsToTrace())) {
                traceException(ex, annotation);
                return handleFallback(pjp, annotation, ex);
            }
            throw ex;
        } finally {
            if (entry != null) {
                entry.exit(1, pjp.getArgs());
            }
        }

四、重点看看SphU.entry的规则校验记录怎么实现的？

跟进上面的那行代码：

entry = SphU.entry(resourceName, resourceType, entryType, pjp.getArgs());

不断跟进会来到entryWithPriority方法：

    private Entry entryWithPriority(ResourceWrapper resourceWrapper, int count, boolean prioritized, Object... args)
        throws BlockException {
        //。。。略

        ProcessorSlot<Object> chain = lookProcessChain(resourceWrapper);
        if (chain == null) {
            return new CtEntry(resourceWrapper, null, context);
        }

        Entry e = new CtEntry(resourceWrapper, chain, context);
        try {
            chain.entry(context, resourceWrapper, null, count, prioritized, args);
        } catch (BlockException e1) {
            e.exit(count, args);
            throw e1;
        } catch (Throwable e1) {
            // This should not happen, unless there are errors existing in Sentinel internal.
            RecordLog.info("Sentinel unexpected exception", e1);
        }
        return e;
    }

主要是lookProcessChain获取一个“链条”，然后执行“链条”的entry的方法（实际上是lookProcessChain取出很多个ProcessorSlot，然后依次执行，责任链设计模式）

4.1 lookProcessChain

    ProcessorSlot<Object> lookProcessChain(ResourceWrapper resourceWrapper) {
       ProcessorSlotChain chain = chainMap.get(resourceWrapper);
       if (chain == null) {
           synchronized (LOCK) {
               chain = chainMap.get(resourceWrapper);
               if (chain == null) {
                   // Entry size limit.
                   if (chainMap.size() >= Constants.MAX_SLOT_CHAIN_SIZE) {
                       return null;
                   }

                   chain = SlotChainProvider.newSlotChain();
                   Map<ResourceWrapper, ProcessorSlotChain> newMap = new HashMap<ResourceWrapper, ProcessorSlotChain>(
                       chainMap.size() + 1);
                   newMap.putAll(chainMap);
                   newMap.put(resourceWrapper, chain);
                   chainMap = newMap;
               }
           }
       }
       return chain;
   }

单例设计模式DCL去创建ProcessorSlotChain，主要是：

chain = SlotChainProvider.newSlotChain();

查看源码会发现是通过建造者模式builder去创建的，而建造者本身builde方法：

    public ProcessorSlotChain build() {
        ProcessorSlotChain chain = new DefaultProcessorSlotChain();
        //加入下面这些slot形成链表，之后调用的时候依次执行
        chain.addLast(new NodeSelectorSlot());
        chain.addLast(new ClusterBuilderSlot());
        chain.addLast(new LogSlot());
        chain.addLast(new StatisticSlot());
        chain.addLast(new SystemSlot());
        chain.addLast(new AuthoritySlot());
        chain.addLast(new FlowSlot());
        chain.addLast(new DegradeSlot());

        return chain;
    }

真相大白，这个执行“执行链条”就是这些各种‘Slot’，核心的是：StatisticSlot（统计相关）、FlowSlot（限流相关）、DegradeSlot（降级相关）。

4.2 chain.entry依次执行各个Slot

回到之前的entryWithPriority方法：

    private Entry entryWithPriority(ResourceWrapper resourceWrapper, int count, boolean prioritized, Object... args)
        throws BlockException {
        //。。。略

        ProcessorSlot<Object> chain = lookProcessChain(resourceWrapper);
        if (chain == null) {
            return new CtEntry(resourceWrapper, null, context);
        }

        Entry e = new CtEntry(resourceWrapper, chain, context);
        try {
            chain.entry(context, resourceWrapper, null, count, prioritized, args);
        } catch (BlockException e1) {
            e.exit(count, args);
            throw e1;
        } catch (Throwable e1) {
            // This should not happen, unless there are errors existing in Sentinel internal.
            RecordLog.info("Sentinel unexpected exception", e1);
        }
        return e;
    }

那么拿到执行链条之后，就是依次执行各个Slot的entry方法了！

五、StatisticSlot（统计相关）

稍后补充todo

六、FlowSlot（限流相关）

稍后补充todo

七、DegradeSlot（降级相关）

稍后补充todo

八、集群限流设计方案-ClusterBuilderSlot

8.1 基本思想

引入了Token Server和Token Client的概念（相当引入了个管理中心）

Token Server：即集群流控服务端，处理来自 Token Client 的请求，根据配置的集群规则判断是否应该发放 token（是否允许通过）。服务端采用的还是滑动窗口。

Token Client：集群流控客户端，就是我们的应用本身，用于向所属 Token Server 通信请求 token（以及统计信息）。集群限流服务端会返回给客户端结果，决定是否限流。

8.2 基本流程

ClusterBuilderSlot用来统计各种集群相关信息啥的，真正判断限流还是在流控FlowRuleChecker类（FlowSlot持有的属性对象）的canPassCheck方法中：

    public boolean canPassCheck(/*@NonNull*/ FlowRule rule, Context context, DefaultNode node, int acquireCount,
                                                    boolean prioritized) {
        String limitApp = rule.getLimitApp();
        if (limitApp == null) {
            return true;
        }
        //判断是集群模式，采用从Token Server获取是否给通过
        if (rule.isClusterMode()) {
            return passClusterCheck(rule, context, node, acquireCount, prioritized);
        }
        return passLocalCheck(rule, context, node, acquireCount, prioritized);
    }

会根据是否是集群模式，采用从Token Server获取是否给通过

8.3 可用性

目前针对token server高可用，sentinel并没有对应的解决方案，默认会降级走本地流控。

8.4 总结

仅靠单机维度去限制的话可能会无法精确地限制总体流量。而集群流控可以精确地控制整个集群的调用总量，结合单机限流兜底，可以更好地发挥流量控制的效果。