H3C S5100-SI Operation Manual page 406

Figure 1-9 Diagram for WRR queuing
WRR queuing schedules all the queues in turn and ensure that all of them can be served
for a certain time by assigning each queue a weight representing a certain amount of
resources. Assume there are eight output queues on the port. WRR assigns queues 7
through 0 the weights w7, w6, w5, w4, w3, w2, w1, and w0.
For example, on a 100 Mbps port, you can configure the weights for WRR queuing to 50, 50,
30, 30, 10, 10, 10, and 10 (corresponding to w7, w6, w5, w4, w3, w2, w1, and w0 in order).
In this way, the queue with the lowest priority can get 5 Mbps (100 Mbps × 1/(5 + 5 + 3 + 3
+ 1 + 1 + 1 + 1)) bandwidth at least, thus avoiding the disadvantage of SP queuing that the
packets in low-priority queues may failed to be served for a long time.
Another advantage of WRR queuing is that though the queues are scheduled in order, the
service time for each queue is not fixed. With WRR, if a queue is empty, the next queue will
be scheduled immediately. In this way, the bandwidth resources are fully utilized.
3) SDWRR
Compared with WRR, SDWRR reduces scheduling delay and smoothes jitter for lower
priority queues.
For example, set the weight values of queue 0 and queue 1 to 5 and 3 respectively. WRR
and SDWRR schedule the queues as follows:
WRR: dequeues the number of packets identical to weight 3 from queue 1 only after the
number of packets identical to weight 5 are dequeued from queue 0. If there is a wide
difference between the weight values of two queues, great delay and jitter will result for
the lower-weight queue.
1-17