H3C Technologies H3C SR8800 - Manual (Page 48)

39
Weighted fair queuing
NOTE:
WFQ is applicable to only SPE cards.
Figure 16 Weighted fair queuing (WFQ)
WFQ is an enhancement to fair queuing (FQ). FQ is designed for fairly sharing network resources,
reducing delay and jitter of traffic. To this end, FQ takes many aspects into consideration to guarantee
fairness among different queues and packets of different sizes during scheduling to
• Prevent packets in low-priority queues from being starved.
• Minimize jitter between packets in each stream by preferentially scheduling short packets in case
both large packets and short packets are waiting in the queues.
Compared with FQ, WFQ takes weights into account when determining the queue scheduling order.
Statistically, WFQ gives high priority traffic more scheduling opportunities than low priority traffic. When
dequeuing packets, WFQ assigns the outgoing interface bandwidth to each traffic flow by precedence.
The higher precedence a traffic flow has, the more bandwidth it gets.
For example, assume that three flows exist on the current interface, with the precedence 1, 2, and 3
respectively. The total bandwidth quota is 6, that is, the sum of all the precedence values. The bandwidth
percentage assigned to each flow is (precedence value of the flow)/total bandwidth quota. Thus, the
bandwidth percentages for the flows are 1/6, 2/6, and 3/6 respectively.
Hardware WFQ provides multiple queues, each with its individual weight.
NOTE:
Queue 0 is a BE queue, queues 1 through 4 are AF queues, queue 5 and queue 6 are EF queues, and
queue 7 is a NC queue. Different types of queues are scheduled by SP, and queues of the same type are
scheduled based on their weights.