3com S7906E 安装指导

 

5-4 

z 

Short packets and long packets are fairly scheduled: if there are both long packets and short 

packets in queues, statistically the short packets should be scheduled preferentially to reduce the 

jitter between packets as a whole.  

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. WFQ 

can automatically classify traffic according to the “session” information of traffic (protocol type, TCP or 

UDP source/destination port numbers, source/destination IP addresses, IP precedence bits in the ToS 

field, and so on), and try to provide as many queues as possible so that each traffic flow can be put into 

these queues to balance the delay of every traffic flow as a whole. When dequeuing packets, WFQ 

assigns the outgoing interface bandwidth to each traffic flow by precedence. The higher precedence 

value a traffic flow has, the more bandwidth it gets.  

Additionally, WFQ can work with the minimum guaranteed bandwidth mechanism. You can configure a 

minimum guaranteed bandwidth for each WFQ queue to guarantee that each WFQ queue is 

guaranteed of the bandwidth when congestion occurs. The assignable bandwidth (assignable 

bandwidth = total bandwidth – the sum of the minimum guaranteed bandwidth for each queue) is 

allocated to queues based on queue priority.  

For example, assume that the total bandwidth of a port is 10 Mbps, and there are five flows on the port 

currently, with the precedence being 0, 1, 2, 3, and 4 and the minimum guaranteed bandwidth being 

128 kbps, 128 kbps, 128 kbps, 64 kbps, and 64 kbps respectively.  

z 

The assignable bandwidth = 10 Mbps – (128 kbps + 128 kbps + 128 kbps + 64 kbps + and 64 kbps) 

= 9.5 Mbps 

z 

The total assignable bandwidth quota is the sum of all the (precedence value + 1)s, that is, 1 + 2 + 

3 + 4 + 5 = 15.  

z 

The bandwidth percentage assigned to each flow is (precedence value of the flow + 1)/total 

assignable bandwidth quota. The bandwidth percentages for the flows are 1/15, 2/15, 3/15, 4/15, 

and 5/15 respectively.  

z 

The bandwidth finally assigned to a queue = the minimum guaranteed bandwidth + the bandwidth 

allocated to the queue from the assignable bandwidth 

Because WFQ can balance delay and jitter among flows when congestion occurs, it is effectively 

applied in some special occasions. For example, WFQ is used for the assured forwarding (AF) services 

of the Resource Reservation Protocol (RSVP). In Generic Traffic Shaping (GTS), WFQ is used to 

schedule buffered packets.  

By assigning some queues on the port to the SP scheduling group and the others to the WRR 

scheduling group (that is, group 1), you implement SP + WRR queue scheduling on the port. Packets in 

the SP scheduling group are scheduled preferentially. When the SP scheduling group is empty, packets 

in the WRR scheduling group are scheduled. Queues in the SP scheduling group are scheduled with 

the SP queue scheduling algorithm. Queues in the WRR scheduling group are scheduled with WRR. 

Complete the following tasks to achieve congestion management: 

Configuring SP Queuing 

Optional 

Optional