26-13
Catalyst 3750 Metro Switch Software Configuration Guide
78-15870-01
Chapter 26 Configuring QoS
Understanding QoS
Because the total ingress bandwidth of all ports can exceed the bandwidth of the internal ring, ingress
queues are located after the packet is classified, policed, and marked and before packets are forwarded
into the switch fabric. Because multiple ingress ports can simultaneously send packets to an egress port
and cause congestion, egress queue-sets are located after the internal ring.
Each port belongs to an egress queue-set, which defines all the characteristics of the four queues per port.
The ES ports also use a hierarchical queueing model in which each packet is assigned to a hierarchical
queue based on the physical interface, VLAN, or class. Traffic destined for an ES port passes through
the queue-set before reaching the hierarchical queues. If congestion occurs in the hierarchical queues
and backs up to the queue-sets, the queue-set configuration controls how traffic is dropped.
Ingress queues use WTD for congestion management. The egress queue-sets also use WTD. For more
information, see the next section.
Ingress queues support SRR in shared mode for scheduling. The egress queue-sets also support SRR in
shared or shaped mode for scheduling. For more information, see the “SRR Shaping and Sharing”
section on page 26-14.
For information about ingress queueing and scheduling, see the “Queueing and Scheduling of Ingress
Queues” section on page 26-15. For information about egress queue-set queueing and scheduling, see
the “Queueing and Scheduling of Egress Queue-Sets” section on page 26-17.
The egress hierarchical queues for ES ports use tail drop or WRED for congestion management. These
ports also use CBWFQ or LLQ for scheduling. For more information, see the “Queueing and Scheduling
of Hierarchical Queues” section on page 26-26.
Weighted Tail Drop
The ingress queues and the egress queue-sets use an enhanced version of the tail-drop
congestion-avoidance mechanism called WTD. WTD manages the queue lengths and provides drop
precedences for different traffic classifications.
As a frame is sent to a particular queue, WTD uses the frame’s assigned QoS label to subject it to
different thresholds. If the threshold is exceeded for that QoS label (the space available in the destination
queue is less than the size of the frame), the switch drops the frame.
Figure 26-6 shows an example of WTD operating on a queue whose size is 1000 frames. Three drop
percentages are configured: 40, 60, and 100 percent. These percentages mean that up to 400 frames can
be queued at the 40-percent threshold, up to 600 frames at the 60-percent threshold, and up to 1000
frames at the 100-percent threshold.
In this example, CoS values 6 and 7 have a greater importance than the other CoS values, and they are
assigned to the 100-percent drop threshold (queue-full state). CoS values 4 and 5 are assigned to the
60-percent threshold, and CoS values 0 to 3 are assigned to the 40-percent threshold.
Suppose the queue is already filled with 600 frames, and a new frame arrives. It contains CoS values 4
and 5 and is subjected to the 60-percent threshold. If this frame is added to the queue, the threshold will
be exceeded, so the switch drops it.
To Next Page
Loading...
Need help?
General
