常用限流算法：令牌桶和漏桶算法解析

漏桶算法

漏桶算法思路很簡單，水（請求）先進入到漏桶裏，漏桶以一定的速度出水，當水流入速度過大會直接溢出，可以看出漏桶算法能強行限制數據的傳輸速率。

令牌桶算法

對於很多應用場景來說，除了要求能夠限制數據的平均傳輸速率外，還要求允許某種程度的突發傳輸。這時候漏桶算法可能就不合適了，令牌桶算法更爲適合。如圖2所示，令牌桶算法的原理是系統會以一個恆定的速度往桶裏放入令牌，而如果請求需要被處理，則需要先從桶裏獲取一個令牌，當桶裏沒有令牌可取時，則拒絕服務。

限流工具類RateLimiter

　　Google開源工具包Guava提供了限流工具類RateLimiter，該類基於令牌桶算法來完成限流，非常易於使用。具體使用可以自行百度。下面是RateLimiter的源碼：

public double acquire() {
        return acquire(1);
    }

 public double acquire(int permits) {
        checkPermits(permits);  //檢查參數是否合法（是否大於0）
        long microsToWait;
        synchronized (mutex) { //應對併發情況需要同步
            microsToWait = reserveNextTicket(permits, readSafeMicros()); //獲得需要等待的時間 
        }
        ticker.sleepMicrosUninterruptibly(microsToWait); //等待，當未達到限制時，microsToWait爲0
        return 1.0 * microsToWait / TimeUnit.SECONDS.toMicros(1L);
    }

private long reserveNextTicket(double requiredPermits, long nowMicros) {
        resync(nowMicros); //補充令牌
        long microsToNextFreeTicket = nextFreeTicketMicros - nowMicros;
        double storedPermitsToSpend = Math.min(requiredPermits, this.storedPermits); //獲取這次請求消耗的令牌數目
        double freshPermits = requiredPermits - storedPermitsToSpend;

        long waitMicros = storedPermitsToWaitTime(this.storedPermits, storedPermitsToSpend)
                + (long) (freshPermits * stableIntervalMicros); 

        this.nextFreeTicketMicros = nextFreeTicketMicros + waitMicros;
        this.storedPermits -= storedPermitsToSpend; // 減去消耗的令牌
        return microsToNextFreeTicket;
    }

private void resync(long nowMicros) {
        // if nextFreeTicket is in the past, resync to now
        if (nowMicros > nextFreeTicketMicros) {
            storedPermits = Math.min(maxPermits,
                    storedPermits + (nowMicros - nextFreeTicketMicros) / stableIntervalMicros);
            nextFreeTicketMicros = nowMicros;
        }
    }