finish rate-limit pattern

resiliency/07_rate_limiting/rate_limiting.go

@@ -12,22 +12,10 @@ With limiting you can controll resource utilization and maintain quality of serv
 Go  supports rate limiting by using goroutines, channels, and tickers.
 */
 
-func rateLimiting(requestQueueSize, allowedBurstCount int) {
+func rateLimiting(requestQueue chan int, duration int64, allowedBurstCount int) {
 
-	requests := make(chan int, requestQueueSize)
-	for i := 1; i <= requestQueueSize; i++ {
-		requests <- i
-	}
-	close(requests)
-
-	limiter := time.Tick(200 * time.Millisecond)
-
-	for req := range requests {
-		<-limiter
-		fmt.Println("request", req, time.Now())
-	}
-
-	//允许的突发数量
+	//根据允许的突发数量,创建ch
+	//只要该队列中有内容，就可以一直取出来，即便ch已经关闭
 	burstyLimiter := make(chan time.Time, allowedBurstCount)
 
 	//初始化允许突发的数量
@@ -37,20 +25,29 @@ func rateLimiting(requestQueueSize, allowedBurstCount int) {
 
 	//控制请求频率的计时器
 	go func() {
-		for t := range time.Tick(200 * time.Millisecond) {
+		for t := range time.Tick(time.Duration(duration) * time.Millisecond) {
 			burstyLimiter <- t
 		}
 	}()
 
-	//请求队列
-	burstyRequests := make(chan int, requestQueueSize)
-	for i := 1; i <= requestQueueSize; i++ {
-		burstyRequests <- i
-	}
-	close(burstyRequests)
-
-	for req := range burstyRequests {
-		<-burstyLimiter
+	for req := range requestQueue {
+		<-burstyLimiter //突发控制器是限流的关键
+		fmt.Println("request", req, time.Now())
+	}
+}
+
+func simpleRateLimiting(duration int64, requestQueueSize int) {
+
+	requests := make(chan int, requestQueueSize)
+	for i := 1; i <= requestQueueSize; i++ {
+		requests <- i
+	}
+	close(requests)
+
+	limiter := time.Tick(time.Duration(duration) * time.Millisecond)
+
+	for req := range requests {
+		<-limiter
 		fmt.Println("request", req, time.Now())
 	}
 }

resiliency/07_rate_limiting/rate_limiting_test.go

@@ -1,3 +1,11 @@
+/*
+ * @Description: https://github.com/crazybber
+ * @Author: Edward
+ * @Date: 2020-05-21 12:14:27
+ * @Last Modified by: Edward
+ * @Last Modified time: 2020-05-21 12:27:56
+ */
+
 package ratelimit
 
 import (
@@ -11,8 +19,23 @@ With limiting you can controll resource utilization and maintain quality of serv
 Go  supports rate limiting by using goroutines, channels, and tickers.
 */
 
+var (
+	requestQueueSize = 10
+)
+
 func TestRateLimiting(t *testing.T) {
 
-	rateLimiting(5, 3)
+	//请求队列
+	burstyRequests := make(chan int, requestQueueSize)
+
+	for i := 1; i <= requestQueueSize; i++ {
+		burstyRequests <- i
+	}
+	close(burstyRequests)
+
+	//对请求进行限流
+
+	//200ms允许一次请求,最多同时3个突发
+	rateLimiting(burstyRequests, 200, 3)
 
 }