Patent attributes
Embodiments of the present invention disclose a network packet loss processing method and apparatus. The method in the present invention includes: comparing a value of a current congestion window of a network with a buffer threshold when a packet loss event is detected, where the buffer threshold is used to indicate a data bearing capability of the entire network, where the data bearing capability is obtained by adding a data bearing capability of a link in the network and a data bearing capability of a buffer of a node in the network; if the value of the current congestion window is smaller than the buffer threshold, not performing backoff processing on the current congestion window; and if the value of the current congestion window is greater than or equal to the buffer threshold, performing backoff processing on the current congestion window.
