A QoS policy is a written document that defines the QoS-related service levels for each service class.
Essentially, the QoS policy includes the documentation of the work performed in the first two steps, plus the definitions of the QoS actions that should be taken in the routers and switches in order to reach the service levels defined in the QoS policy document.
In order for a QoS policy to actually provide a particular level of service, the policy defines actions to be taken on the packets inside a service class. These actions cause some change in the bandwidth, delay, jitter, and/or loss characteristics of those packets. For instance, a QoS policy might call for a queuing tool to give a guaranteed amount of bandwidth to a class of traffic. Another example might be to more aggressively discard one service class’s packets, in order to reduce packet loss for another service class. A QoS policy might set a limit to the amount of bandwidth allowed for a service class, in order to prevent that service class from taking more bandwidth than it should.
While these steps may seem somewhat intuitive, once you read more about the QoS tools themselves, particular those that use the Modular QoS CLI (MQC), you will see that this 3-step process does provide a good basic framework with which to attack QoS efforts.
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 will help solidify some key facts. For any of you doing your final prep before the exam, these tables and figures will be a convenient way to review the day before the exam.
Table 1-27 outlines some of the behaviors seen when no QoS is applied in a network.
As shown in Figure 1-36, with compression, if a ratio of 2:1 is achieved, the 80-kbps flow will only require 40 kbps in order to be sent across the link—effectively doubling the bandwidth capacity of the link.
Table 1-27 Traffic Behavior with No QoS
Type of Traffic Behavior Without QoS Voice Voice is hard to understand.
Voice breaks up, sounds choppy.
Delays make interacting difficult; callers do not know when other party has finished talking.
Calls are disconnected.
Video Picture displays erratically; jerky movements.
Audio not in sync with video.
Movement slows down.
Data Data arrives after it is no longer useful.
Customer waiting for customer care agent, who waits for a screen to display.
Erratic response times frustrate users, who may give up or try later.
Figure 1-36 With a 2:1 Compression Ratio Versus No Compression
Figure 1-37 shows a two-queue system where the first queue gets 25 percent of the bandwidth on the link, and the second queue gets 75 percent of the bandwidth.
Figure 1-37 Bandwidth Reservation Using Queuing
Figure 1-38 shows two contrasting examples of serialization and propagation delay.
FIFO Queue Router 1
Offered Load:
80 kbps Compress
Tx Queue
R2 S0
40 kbps Sent 64 kbps
FIFO Queue Router 1
Offered Load:
80 kbps
Tx Queue
R2 S0
64 kbps Sent, Rest Queued 64 kbps
Larger Queue Due to Congestion
4 X 1500 Byte Packets
R1
Output Queue 1 1 2 3
Output Queue 2 R2 4
25%
Bandwidth
75%
Bandwidth
Figure 1-38 Serialization and Propagation Delay for Selected Packet and Link Lengths
Figure 1-39 lists the queuing, serialization, and propagation delays experienced by data, voice, and video traffic.
Figure 1-39 Delay Components: Three Components, Single Router (R1)
Figure 1-40 depicts LFI operation.
Figure 1-40 Link Fragmentation and Interleaving
R1 R2 R3
SW1 Hannah
1000 km
56 kbps 128 kbps
10 km
T1 SW2
T3
SW3
SW4 Server 1
Serialization: 125-Byte Packet: 17.86 ms Serialization: 1250-Byte Packet: 178.6 ms Propagation: Either Size Packet: 4.8 ms
Serialization: 125-Byte Packet: .65 ms Serialization: 1250-Byte Packet: 6.5 ms Propagation: Either Size Packet: .048 ms
Propagation Delay: 4.8 ms R2 4 X 1500
Byte Packets
R1
FIFO Output Queue 1 2 3
4 1001101110101011 Serialization Delay: 214 ms
Packet 1: 1500 Bytes, Arrives First Packet 2: 200 Bytes, Delay Sensitive, Arrives Second
R1
Output Queue 2 R2
Output Queue 1: 3 Fragments of Packet #1 Shown
P1 F3
P1 F2
P1
F1 P1
F2 2 P1 F1 P1 F3 2
Output Queue 1
Legend: Px Fy Means Packet Number x, Fragment Number y
Figure 1-41 shows the jitter experienced by three packets as part of a voice call between phones at extension 301 and 201.
Figure 1-41 Jitter Example
Figure 1-42 outlines the format of an IP packet using RTP.
Figure 1-42 IP Packet for Voice Call—RTP
R1
IP
R2 R3
IP
SW1 Hannah
201
s0 s0 s1 T1 s0
FA0
SW2
301
Server 1
RTP RTP RTP
20 20
RTP RTP RTP
20 30
IP UDP RTP Voice Payload
20 Bytes 8 Bytes 12 Bytes Variable
Port Ranges:
16384 - 32767 (Even Ports)
Popular Values:
G.711: 160 Bytes G729a: 20 Bytes
Table 1-28 lists the bandwidth requirements when using one of two codecs, with varying types of data link protocols.
*Layer 3 bandwidth consumption refers to the amount of bandwidth consumed by the Layer 3 header through the data (payload) portion of the packet.
Figure 1-43 shows an example of delay concepts, with sample delay values shown. When the delay is negligible, the delay is just listed as zero.
Figure 1-43 Example Network with Various Delay Components shown: Left-to-Right Direction
Table 1-28 Updated Bandwidth Requirements for Various Types of Voice Calls
Bandwidth Consumption, Including L2 Overhead
Layer 3 Bandwidth Consumption
*
802.1Q Ethernet (32 Bytes of L2 Overhead)
PPP (9 Bytes of L2 Overhead)
MLP (13 Bytes of L2 Overhead)
Frame-Relay (8 Bytes of L2 Overhead)
ATM (Variable Bytes of L2 Overhead, Depending on Cell-Padding Requirements)
G.711 at 50 pps 80 kbps 93 kbps 84 kbps 86 kbps 84 kbps 106 kbps
G.711 at 33 pps 75 kbps 83 kbps 77 kbps 78 kbps 77 kbps 84 kbps
G.729A at 50 pps 24 kbps 37 kbps 28 kbps 30 kbps 28 kbps 43 kbps
G.729A at 33 pps 19 kbps 27 kbps 21 kbps 22 kbps 21 kbps 28 kbps
R1
IP
R2 R3
IP
SW1 Hannah
201
s0 s0 s1 T1 s0/0
FA0/0 SW2
301
Server 1
Forwarding: 0 Queuing: 15 Serialization: 9 Propagation: .5
Network: 50 (Note: Do Not Count R2 Serialization Here and at R2!) Forwarding: 0
Queuing: 0 Serialization: 0
Forwarding: 0 Queuing: 15 Serialization: 4 Propagation: .5
Forwarding: 0 Queuing: 0 Serialization: 0 Propagation: 0 Delays for Packets Flowing Left-to-Right: Total Delay: 94 ms
Table 1-29 outlines the suggested delay budgets.
All the delay components for a voice call are summarized in the example in Figure 1-44.
Figure 1-44 Complete End-to-End Voice Delay Example
Table 1-30 lists the different delay components and whether they are variable.
Table 1-29 One-Way Delay Budget Guidelines 1-Way Delay (in ms) Description
0–150 ITU G.114 recommended acceptable range
0–200 Cisco’s recommended acceptable range
150–400 ITU G.114’s recommended range for degraded service 400+ ITU G.114’s range of unacceptable delay in all cases
Table 1-30 Delay Components, Variable and Fixed Delay
Component
Fixed or
Variable Comments QoS Tools That Can Help
Codec Fixed Varies slightly based on codec and processing load; considered fixed in course books (and probably on exams). Typically around 10 ms.
None.
R1
IP
R2 R3
IP
SW1 Hannah
201
s0 s0 s1 T1 s0/0
FA0/0 SW2
301
Server 1
Forwarding: 0 Queuing: 15 Serialization: 9 Propagation: .5
Network: 50 (Note: Do Not Count R2 Serialization Here and at R2!) Forwarding: 0
Queuing: 0 Serialization: 0
Forwarding: 0 Queuing: 15 Serialization: 4 Propagation: .5
Forwarding: 0 Queuing: 0 Serialization: 0 Propagation: 0 Delays for Packets Flowing Left-to-Right: Total Delay: 164 ms
Codec: 10 Packetization: 20
De-jitter: 40 ms De-jitter: 40 ms
Table 1-31 summarizes the QoS requirements of data, in comparison to voice and video.
Delay Component
Fixed or
Variable Comments QoS Tools That Can Help
Packetization Fixed Some codecs require a 30-ms payload, but packetization delay does not vary for a single codec.
Typically 20 ms, including when using G.711 and G.729.
None.
Propagation Variable Varies based on length of circuit.
About 5 ms/100 km
Move your facilities to the same town.
Queuing Variable This is the most controllable delay component for packet voice
Queuing features, particularly those with a priority-queuing feature.
Serialization Fixed It is fixed for voice packets, because all voice packets are of equal length.
It is variable based on packet size for all packets.
Fragmentation and compression.
Network Variable Least controllable variable component.
Shaping, fragmentation, designs mindful of reducing delay.
De-jitter buffer (initial playout delay)
Variable This component is variable because it can be configured for a different value. However, that value, once configured, remains fixed for all calls until another value is configured. In other words, the initial playout delay does not dynamically vary.
Configurable playout delay in IOS gateways; not
configurable in IP Phones.
Table 1-31 Comparing Voice, Video, and Data QoS Requirements
Bandwidth Delay Jitter Loss
Voice Payload Low to Medium Low Low Low
Video Payload Interactive (2Way)
Medium Low Low Low
Video Payload Streaming (1Way)
Medium to High High High Low
Video Signaling Low Low Medium Medium
Voice Signaling Low Low Medium Medium
Table 1-30 Delay Components, Variable and Fixed (Continued)
continues
Bandwidth Delay Jitter Loss Data: Interactive,
Mission Critical
Low to Medium Low to Medium
Low to Medium
Medium to high Data: Not Interactive,
Mission Critical
Variable, typically high
High High Medium
Data: Interactive, Not Critical
Variable, typical medium
High High Medium
Data: Not Interactive, Not Critical
Variable, typically high
High High High
Table 1-31 Comparing Voice, Video, and Data QoS Requirements (Continued)
Q&A
As mentioned in the Introduction, you have two choices for review questions. The questions that follow next give you a more difficult challenge than the exam itself by using an open-ended question format. By reviewing now with this more difficult question format, you can exercise your memory better, and prove your conceptual and factual knowledge of this chapter. You can find the answers to these questions in Appendix A.
The second option for practice questions is to use the CD-ROM included with this book. It includes a testing engine and more than 200 multiple-choice questions. You should use this CD-ROM nearer to the end of your preparation, for practice with the actual exam format.
1. List the four traffic characteristics that QoS tools can affect.
2. Describe some of the characteristics of voice traffic when no QoS is applied in a network.
3. Describe some of the characteristics of video traffic when no QoS is applied in a network.
4. Describe some of the characteristics of data traffic when no QoS is applied in a network.
5. Interpret the meaning of the phrase, “QoS is both ‘managed fairness,’ and at the same time
‘managed unfairness’.”
6. Define bandwidth. Compare and contrast bandwidth concepts over point-to-point links versus Frame Relay.
7. Compare and contrast bandwidth and clock rate in relation to usage for QoS.
8. List the QoS tool types that affect bandwidth, and give a brief explanation of why each tool can affect bandwidth.
9. Define delay, compare/contrast one-way and round-trip delay, and characterize the types of packets for which one-way delay is important.
10. List the categories of delay that could be experienced by all three types of traffic: data, voice, and video.
11. Define, compare, and contrast serialization and propagation delay.
12. Define network delay.
13. List the QoS tool types that affect delay and give a brief explanation of why each tool can affect delay.
14. Define jitter. Give an example that shows a packet without jitter, followed by a packet with jitter.
15. List the QoS tool types that affect jitter and give a brief explanation of why each tool can affect jitter.
16. Define packet loss and describe the primary reason for loss for which QoS tools can help.
17. List the QoS tool types that affect loss and give a brief explanation of why each tool can affect loss.
18. Describe the contents of an IP packet carrying the payload for a G.729 VoIP call.
19. Describe the amount of bandwidth required for G.711 and G.729 VoIP calls, ignoring data-link header/trailer overhead.
20. List the delay components that voice calls experience, but which data-only flows do not experience.
21. Define the meaning of the term “packetization delay” in relation to a voice call.
22. List the different one-way delay budgets as suggested by Cisco and the ITU.
23. Define the term “codec delay” and discuss the two components when using a G.729 codec.
24. Describe the affects of a single lost packet versus two consecutive lost packets, for a G.729 voice call.
25. Describe a typical video payload flow in terms of packet sizes and packet rates.
26. Discuss the delay requirements of video traffic.
27. List the basic differences between TCP and UDP traffic.
28. Contrast the QoS characteristics needed by interactive data applications, as compared to the QoS needs of voice payload flows.
29. What are the three steps suggested in this chapter for planning QoS policy implementation?
30. The chapter provides a suggested process for planning and implementing QoS policies. The first step involves identifying traffic classes and requirements. That step lists two specific types of audits that should be performed in this step. List and give a brief description of the two audit steps.
31. Early in the chapter, a couple of different definitions of QoS were supplied. Paraphrase your own general definition of the term “QoS”.
32. What is the purpose of service classes when implementing a QoS policy?
33. What are the three steps suggested in this chapter for planning QoS policy implementation?
34. The chapter provides a suggested process for planning and implementing QoS policies. The first step involves identifying traffic classes and requirements. That step lists two specific types of audits that should be performed in this step. List and give a brief description of the two audit steps.
35. Early in the chapter, a couple of different definitions of QoS were supplied. Paraphrase your own general definition of the term “QoS”.
36. What is the purpose of service classes when implementing a QoS policy?
QoS Exam Topics
This chapter covers the following exam topics specific to the QOS exam:
■ List and explain the models for providing Quality of Service on a network
■ Explain the purpose and function of the DiffServ model
■ Describe the basic format of and explain the purpose of the DSCP field in the IP header
■ Define and explain the different per hop behaviors used in DSCP
■ Explain the interoperability between DSCP-based and IP-precedence-based devices in a network
■ Given a list of QoS actions, correctly match the QoS actions to mechanisms for implementing QoS and identify where in a network the different QoS mechanisms are commonly used
2
QoS Tools and Architectures
To build a house, you need tools, you need materials, you need labor, and you need architectural plans. To build a network using quality of service (QoS), you need tools, labor, and an architecture.
This chapter lists the various IOS QoS tools and explains the two predominant QoS architectures: integrated services (IntServ) and differentiated services (DiffServ).
Chapter 1, “QoS Overview,” covered various types of QoS tools. There are several different categories of QoS tools in Cisco IOS Software. This chapter begins by listing the tools in each category—at least the ones covered on the QOS exam. This first section also includes a brief introduction to concepts behind each type of tool. All the tools listed here get further treatment in later chapters of the book.
As a tool for learning, the second section of this chapter explains the basics of flow-based QoS tools and Class-Based QoS tools. Taking a few minutes to think about these concepts, and why they make sense, is useful before looking at the two formalized QoS models — namely DiffServ and IntServ.
Next, this chapter then examines the DiffServ architecture in detail. DiffServ attempts to provide Internet-scale QoS, which is a lofty goal indeed! DiffServ uses a Class-Based approach to differentiate between packets, which scales somewhat better than its predecessor, IntServ.
Whether DiffServ succeeds in this goal remains to be seen; however, many of the concepts can be helpful with any QoS implementation.
Finally, the chapter ends with a short discussion on IntServ and Best Effort architectures.
IntServ uses the Resource Reservation Protocol (RSVP) to reserve bandwidth for individual flows in the network. Best Effort is actually just a plan in which the network makes no real effort to give one type of packet better service than any other.
This chapter concludes the introductory materials in this book; the remainder of this book delves into the details of the various QoS tools.
“Do I Know This Already?” Quiz
The purpose of the “Do I Know This Already?” quiz is to help you decide whether you need to read the entire chapter. If you already intend to read the entire chapter, you do not necessarily need to answer these questions now.
The 12-question quiz, derived from the major sections in the “Foundation Topics” section of this chapter, helps you determine how to spend your limited study time.
Table 2-1 outlines the major topics discussed in this chapter and the “Do I Know This Already?”
quiz questions that correspond to those topics.
You can find the answers to the “Do I Know This Already?” quiz in Appendix A, “Answers to the
‘Do I Know This Already?’ Quizzes and Q&A Sections.” The suggested choices for your next step are as follows:
■ 10 or less overall score—Read the entire chapter. This includes the “Foundation Topics,” the
“Foundation Summary,” and the “Q&A” sections.
■ 11 or 12 overall score—If you want more review on these topics, skip to the “Foundation Summary” section and then go to the “Q&A” section. Otherwise, proceed to the next chapter.
Table 2-1 “Do I Know This Already?” Foundation Topics Section-to-Question Mapping
Foundation Topics Section Covering These Questions Questions Score
QoS Tools 1–4
Classifying Using Flows or Service Classes 5–6
The Differentiated Services QoS Model 7–11
The Integrated Services QoS Model 12
Total
CAUTION The goal of self-assessment is to gauge your mastery of the topics in this chapter.
If you do not know the answer to a question or are only partially sure of the answer, mark this question wrong for purposes of the self-assessment. Giving yourself credit for an answer you correctly guess skews your self-assessment results and might provide you with a false sense of security.
QoS Tools Questions
1. Which of the following are not Queuing tools?
a. LLQ
b. CBPQ
c. CBWFQ
d. CBCQ
e. WRR
2. Which of the following tools monitors the rate at which bits are sent out an interface?
a. LLQ
b. CB Shaping
c. WRED
d. CB Policing
e. MLP LFI
3. Which of the following tools can mark IP packet’s DSCP field?
a. CB Marking
b. WRED
c. CB Policing
d. MLP LFI
e. NBAR
4. Which of the following tools chooses to discard packets even though the router either has memory to queue the packets, or available physical bandwidth to send the packets?
a. CB Marking
b. WRED
c. CB Policing
d. MLP LFI
e. NBAR
f. ECN
Classifying Using Flows or Service Classes Questions
5. Which of the following are not used to identify a flow?
a. ECN bits
b. Source port
c. Layer 4 protocol type
d. Destination IP address
e. TCP acknowledgment number
6. Which of the following are likely places at which to mark packets in a network using good QoS design practices?
a. In client PCs
b. As close to the client as possible without allowing end users to mark packets
c. As packets enter an SP from a customer network
d. As packets enter the LAN switch to which the destination host is attached
The Differentiated Services QoS Model Questions
7. What does DSCP stand for?
a. Diffserv Static Code Point
b. Diffserv Standardized Configuration Process
c. Differentiated Services Code Point
d. Differentiated Services Configuration Point
8. According to the DiffServ, which PHB defines a set of three DSCPs in each service class, with different drop characteristics for each of the three DSCP values?
a. Expedited Forwarding
b. Class Selector
c. Assured Forwarding
d. Multi-class-multi-drop
9. Which of the following is true about the location of DSCP?
a. High order 6 bits of ToS byte
b. Low order 6 bits of ToS byte
c. Middle 6 bits of ToS byte
d. High order 5 bits in ToS byte