| 摘要 |
This is a method for use in detecting and measuring microcongestion on links where QoS is utilized to provide a prioritized packet delivery service. Microcongestion is a transient condition that occurs over extremely short periods of time and is generally invisible to the end user. However, the ability to monitor the frequency and severity of microcongestion can help identify link capacity issues with far greater accuracy than standard passive measurements and at a much earlier stage than traditional active measurements. This capability can be particularly valuable on very-high-speed links where even small periods of perceptible congestion can represent a significant number of queued packets.
|
