/* * @Description: https://github.com/crazybber * @Author: Edward * @Date: 2020-05-22 12:42:34 * @Last Modified by: Edward * @Last Modified time: 2020-05-22 17:21:40 */ package circuit import ( "context" "time" ) //////////////////////////////// //way 1 简单的函数式断路器 // 一个func实例作为一个Breaker 允许多个worker(即goroutine)同时访问 // 当前简单场景下,只考虑单个worker情况下的连续请求 // 当前的设计思路,非常简单: // 1、不考虑三种状态之间的转换,只靠两种状态,电路打开与电路关闭 // 2、当累计失败到达一定失败次数就端开请求(打开电路),并延迟一定的时间再允许请求 //////////////////////////////// //states of CircuitBreaker //states: closed --->open ---> half open --> closed const ( UnknownState OperationState = iota FailureState SuccessState ) type simpleCounter struct { lastOpResult OperationState lastActivity time.Time ConsecutiveSuccesses uint32 ConsecutiveFailures uint32 } func (c *simpleCounter) LastActivity() time.Time { return c.lastActivity } func (c *simpleCounter) Reset() { ct := &simpleCounter{} ct.lastActivity = c.lastActivity ct.lastOpResult = UnknownState c = ct } //Count the failure and success func (c *simpleCounter) Count(lastState OperationState) { switch lastState { case FailureState: c.ConsecutiveFailures++ case SuccessState: c.ConsecutiveSuccesses++ } c.lastActivity = time.Now() //更新活动时间 c.lastOpResult = lastState //handle status change } //Circuit of action stream,this is actually to do something. //Circuit hold the really action type Circuit func(context.Context) error //失败达到阈值后,过两秒重试 var canRetry = func(cnt simpleCounter, failureThreshold uint32) bool { backoffLevel := cnt.ConsecutiveFailures - failureThreshold // Calculates when should the circuit breaker resume propagating requests // to the service shouldRetryAt := cnt.LastActivity().Add(time.Second << backoffLevel) return time.Now().After(shouldRetryAt) } //Breaker return a closure wrapper to hold Circuit Request func Breaker(c Circuit, failureThreshold uint32) Circuit { //闭包内部的全局计数器 和状态标志 cnt := simpleCounter{} //ctx can be used hold parameters return func(ctx context.Context) error { //阻止请求 if cnt.ConsecutiveFailures >= failureThreshold { if !canRetry(cnt, failureThreshold) { // Fails fast instead of propagating requests to the circuit since // not enough time has passed since the last failure to retry return ErrServiceUnavailable } //reset mark for failures cnt.ConsecutiveFailures = 0 } // Unless the failure threshold is exceeded the wrapped service mimics the // old behavior and the difference in behavior is seen after consecutive failures if err := c(ctx); err != nil { //连续失败会增大backoff 时间 cnt.Count(FailureState) return err } cnt.Count(SuccessState) return nil } }