Foundation Summary

The "Foundation Summary" is a collection of tables and figures that provide a convenient review of many key concepts in this chapter. For those of you already comfortable with the topics in this chapter, this summary could help you recall a few details. For those of you who just read this chapter, this review should help solidify some key facts. For any of you doing your final preparation before the exam, these tables and figures are a convenient way to review the day before the exam.

The following list summarizes the key points about TX Rings and TX Queues in relation to their effect on queuing:

• The TX Queue/TX Ring always performs FIFO scheduling, and cannot be changed.

• The TX Queue/TX Ring uses a single queue, per interface.

• IOS shortens the interface TX Queue/TX Ring automatically when an output queuing method is configured

• You can configure the TX Ring/TX Queue length to a different value.

Figure 5-22 shows how Hardware Queues affect queuing. With queuing configured with two queues, seven packets arrive, numbered in the order in which they arrive. The output queuing configuration specifies that the first two packets (1 and 2) should be placed into Queue 2, and the next four packets (numbered 3 through 6) should be placed into Queue 1.

To delay the traffic, traffic shaping places the packets into the queue associated with the subinterface or DLCI and drains the traffic from the shaping queue at the shaped rate. Figure 5-23 shows the structure of the queues on a subinterface, interface, and the TX Queue, when shaping is enabled.

Flow-Based WFQ, or simply WFQ, classifies traffic into flows. Flows are identified by at least five items in an IP packet.

• Source IP address

• Destination IP address

• Transport layer protocol (TCP or UDP) as defined by the IP Protocol header field

• TCP or UDP source port

• TCP or UDP destination port

• IP Precedence

WFQ calculates the sequence number (SN) (also called Finish Time (FT)) before adding a packet to its associated queue. The formula for calculating the SN for a packet is as follows:

Table 5-12 lists the weight values used by WFQ as of 12.0(5)T/12.1.

Table 5-12. Weight Values Used by WFQ

Precedence	After 12.0(5)T/12.1
0	32384
1	16192
2	10794
3	8096
4	6476
5	5397
6	4626
7	4048

WFQ discards some packet when a queue's congestive discard threshold (CDT) has been reached. To appreciate how the CDT is used, examine Figure 5-24.

Table 5-13 summarizes some of the key features of WFQ.

Table 5-13. WFQ Functions and Features

WFQ Feature	Explanation
Classification	Classifies without configuration, based on source/destination IP address/ port, protocol type (TCP|UDP), and ToS.
Drop policy	Modified tail drop.
Number of queues	4096.
Maximum queue length	Congestive discard threshold per queue (max 4096), with an overall limit based on the hold queue for all queues (max 4096).
Scheduling inside a single queue	FIFO.
Scheduling among all queues	Serves lowest sequence number (SN). The SN is assigned when the packet is placed into the queue, as a function of length and precedence.

Table 5-14 summarizes some of the key features of CBWFQ.

Table 5-14. CBWFQ Functions and Features

CBWFQ Feature	Description
Classification	Classifies based on anything that MQC commands can match, just like CB marking. Includes all extended IP ACL fields, NBAR, incoming interface, CoS, precedence, DSCP, source/destination MAC, MPLS Experimental, QoS group, and RTP port numbers
Drop policy	Tail drop or WRED, configurable per queue.
Number of queues	64.
Maximum queue length	Varies per router model and memory.
Scheduling inside a single queue	FIFO on 63 queues; FIFO or WFQ on class-default queue[*].
Scheduling among all queues	Algorithm is not published. The result of the scheduler provides a percentage guaranteed bandwidth to each queue.

^[*] Except on 7500 series, where you can use FIFO or WFQ in all the CBWFQ queues.

To prevent LLQ from having the same problem as PQ, where packets in the highest-priority queue could dominate, LLQ's scheduler actually works as shown in Figure 5-25.

Table 5-15 lists a few of the more important points about these queuing tools, with comments about their support of each point.

Table 5-15. Comparisons of WFQ, CBWFQ, and LLQ

Concept	WFQ	CBWFQ	LLQ
Requires complex classification configuration	No	Yes	Yes
Uses MQC	No	Yes	Yes
Prefers low volume, high precedence flows	Yes	Not flow based	Not flow based
Experiences problems with large numbers of flows	Yes	No[*]	No
Can reserve bandwidth per queue	No	Yes	Yes
Provide low delay, low jitter queuing	No	No	yes

^[*] With WFQ enabled inside the a CBWFQ class-default queue, that class can experience problems with large numbers of flows.

For Further Reading

This book attempts to cover the breadth and depth of QoS as covered on the QoS exam (642-642). However, you may want to read more about topics in this chapter, or other classification and marking topics.

For more on the topics in this chapter:

• Cisco IOS 12.3 QOS Configuration Guide (http://www.cisco.com/univercd/cc/td/doc/product/software/ios123/123cgcr/qos_vcg.htm)

For more on other Classification and Marking topics:

• Appendix B, "Additional QoS Reference Materials" (found on the book's accompanying CD-ROM):

  Priority Queuing (PQ)

  Custom Queuing (CQ)

  Modified Deficit Round Robin (MDRR)

For design related guidance:

• Cisco's document "Cisco AVVID Network Infrastructure Enterprise Quality of Service Design" document at http://cisco.com/application/pdf/en/us/guest/netsol/ns17/c649/ccmigration_09186a00800d67ed.pdf