Available:*
Shelf Number | Material Type | Copy | Shelf Location | Status |
---|---|---|---|---|
004.6 22 | 1:E-BOOK | 1 | 1:ONLINE | Searching... Unknown |
Bound With These Titles
On Order
Summary
Summary
QoS, short for "quality of service," is one of the most important goals a network designer or administrator will have. Ensuring that the network runs at optimal precision with data remaining accurate, traveling fast, and to the correct user are the main objectives of QoS. The various media that fly across the network including voice, video, and data have different idiosyncrasies that try the dimensions of the network. This malleable network architecture poses an always moving potential problem for the network professional.
The authors have provided a comprehensive treatise on this subject. They have included topics such as traffic engineering, capacity planning, and admission control. This book provides real world case studies of QoS in multiservice networks. These case studies remove the mystery behind QoS by illustrating the how, what, and why of implementing QoS within networks. Readers will be able to learn from the successes and failures of these actual working designs and configurations.
Table of Contents
Preface | p. xiii |
Acknowledgments | p. xxi |
About the authors | p. xxiii |
1 QOS Requirements and Service Level Agreements | p. 1 |
1.1 Introduction | p. 1 |
1.2 SLA Metrics | p. 4 |
1.2.1 Network Delay | p. 4 |
1.2.1.1 Propagation Delay | p. 5 |
1.2.1.2 Switching Delay | p. 6 |
1.2.1.3 Scheduling Delay | p. 6 |
1.2.1.4 Serialization Delay | p. 6 |
1.2.2 Delay-jitter | p. 8 |
1.2.3 Packet Loss | p. 9 |
1.2.4 Bandwidth and Throughput | p. 12 |
1.2.4.1 Layer 2 Overheads | p. 13 |
1.2.4.2 VPN Hose and Pipe Models | p. 16 |
1.2.5 Per Flow Sequence Preservation | p. 18 |
1.2.6 Availability | p. 20 |
1.2.6.1 Network Availability | p. 20 |
1.2.6.2 Service Availability | p. 21 |
1.2.7 Quality of Experience | p. 22 |
1.2.7.1 Voice | p. 23 |
1.2.7.2 Video | p. 24 |
1.2.7.3 On-line Gaming | p. 24 |
1.3 Application SLA Requirements | p. 24 |
1.3.1 Voice over IP | p. 26 |
1.3.1.1 VoIP: Impact of Delay | p. 29 |
1.3.1.2 VoIP: Impact of Delay-jitter | p. 31 |
1.3.1.3 VoIP: Impact of Loss | p. 33 |
1.3.1.4 VoIP: Impact of Throughput | p. 36 |
1.3.1.5 VoIP: Impact of Packet Re-ordering | p. 37 |
1.3.2 Video | p. 38 |
1.3.2.1 Video Streaming | p. 38 |
1.3.2.2 Video Conferencing | p. 57 |
1.3.3 Data Applications | p. 58 |
1.3.3.1 Throughput Focussed TCP Applications | p. 59 |
1.3.3.2 Interactive Data Applications | p. 70 |
1.3.3.3 On-line Gaming | p. 74 |
1.4 Marketed SLAs versus Engineered SLAs | p. 16 |
1.4.1 End-to-End SLAs vs Segmented SLAs | p. 77 |
1.4.2 Inter-provider SLAs | p. 77 |
1.5 Intserv and Diffserv SLAs | p. 78 |
References | p. 79 |
2 Introduction to QOS Mechanics and Architectures | p. 87 |
2.1 What is Quality of Service? | p. 87 |
2.1.1 Quality of Service vs Class of Service or Type of Service? | p. 88 |
2.1.2 Best-effort Service | p. 89 |
2.1.3 The Timeframes that Matter for QOS | p. 90 |
2.1.4 Why IP QOS? | p. 91 |
2.1.5 The QOS Toolset | p. 91 |
2.2 Data Plane QOS Mechanisms | p. 94 |
2.2.1 Classification | p. 94 |
2.2.1.1 Implicit Classification | p. 95 |
2.2.1.2 Complex Classification | p. 95 |
2.2.1.3 Deep Packet Inspection/Stateful Inspection | p. 96 |
2.2.1.4 Simple Classification | p. 96 |
2.2.2 Marking | p. 99 |
2.2.3 Policing and Metering | p. 100 |
2.2.3.1 RFC 2697: Single Rate Three Color Marker | p. 102 |
2.2.3.2 RFC 2698: Two Rate Three Color Marker | p. 106 |
2.2.3.3 Color-aware Policers | p. 108 |
2.2.3.4 Metering | p. 111 |
2.2.4 Queuing, Scheduling, Shaping, and Dropping | p. 112 |
2.2.4.1 Queuing and Scheduling | p. 112 |
2.2.4.2 Dropping | p. 128 |
2.2.4.3 Shaping | p. 137 |
2.2.5 Link Fragmentation and Interleaving | p. 140 |
2.3 IP QOS Architectures | p. 141 |
2.3.1 A Short History of IP Quality of Service | p. 141 |
2.3.2 Type of Service/IP Precedence | p. 142 |
2.3.2.1 IP Precedence | p. 144 |
2.3.2.2 Type of Service | p. 145 |
2.3.2.3 IPv6 Traffic Class Octet | p. 147 |
2.3.3 Integrated Services Architecture | p. 147 |
2.3.4 Differentiated Services Architecture | p. 147 |
2.3.4.1 DS Field | p. 150 |
2.3.4.2 Per-Hop Behaviors | p. 154 |
2.3.4.3 Per-Domain Behaviors | p. 159 |
2.3.4.4 Explicit Congestion Notification | p. 160 |
2.3.4.5 Diffserv Tunneling Models | p. 165 |
2.3.5 IPv6 QOS Architectures | p. 170 |
2.3.6 MPLS QOS Architectures | p. 171 |
2.3.6.1 MPLS and Intserv/RSVP | p. 172 |
2.3.6.2 MPLS and Diffserv | p. 173 |
2.3.7 IP Multicast and QOS | p. 181 |
2.4 Typical Router QOS Implementations in Practice | p. 183 |
2.5 Layer 2 QOS | p. 189 |
2.5.1 ATM | p. 190 |
2.5.1.1 Mapping Diffserv to ATM QOS | p. 193 |
2.5.2 Frame-relay | p. 194 |
2.5.3 Ethernet | p. 196 |
2.6 Complementary Technologies | p. 197 |
2.7 Where QOS cannot make a difference | p. 198 |
References | p. 199 |
Appendix 2.A Precedence, TOS, and DSCP Conversion | p. 204 |
2.A.1 Notation | p. 204 |
2.A.2 Conversion | p. 205 |
3 Deploying Diffserv | p. 209 |
3.1 Introduction | p. 209 |
3.2 Deploying Diffserv at the Network Edge | p. 211 |
3.2.1 Why is the Edge Key for Tight SLA Services? | p. 211 |
3.2.2 Edge Diffserv Case Study | p. 212 |
3.2.2.1 SLA Specification | p. 212 |
3.2.2.2 Diffserv Meta-Language | p. 218 |
3.2.2.3 High-speed Edge Design | p. 218 |
3.2.2.4 Design Variations | p. 225 |
3.2.2.5 Edge SLA Summary | p. 241 |
3.2.2.6 How Many Classes are Enough? | p. 241 |
3.2.2.7 What Marking Scheme to Use? | p. 244 |
3.2.2.8 VoIP - How Much is Enough at the Edge? | p. 245 |
3.3 Deploying Diffserv in the Network Backbone | p. 249 |
3.3.1 Is Diffserv Needed in the Backbone? | p. 249 |
3.3.2 Core Case Study | p. 253 |
3.3.2.1 Core Classes of Service and SLA Specification | p. 253 |
3.3.2.2 "Prioritized" Diffserv Core Model | p. 254 |
3.3.2.3 Detailed Core Design | p. 256 |
3.3.2.4 Design Variations | p. 261 |
3.3.2.5 Core-marking Scheme | p. 263 |
3.4 Tuning (W)RED | p. 268 |
3.4.1 Tuning the Exponential Weighting Constant | p. 269 |
3.4.2 Tuning Minth and Maxth | p. 270 |
3.4.3 Mark Probability Denominator | p. 271 |
3.4.4 In- and Out-of-contract | p. 271 |
References | p. 272 |
4 Capacity Admission Control | p. 275 |
4.1 Introduction | p. 275 |
4.1.1 When is Admission Control Needed? | p. 277 |
4.1.2 A Taxonomy for Admission Control | p. 282 |
4.1.3 What Information is Needed for Admission Control? | p. 285 |
4.1.4 Parameterized or Measurements-based Algorithms | p. 286 |
4.1.4.1 Parameterized Algorithms | p. 286 |
4.1.4.2 Measurement-based Algorithms | p. 288 |
4.2 Topology-unaware Off-path CAC | p. 290 |
4.3 Topology-aware Off-path CAC: "Bandwidth Manager" | p. 292 |
4.3.1 Example Bandwidth Manager Method of Operation: Next Generation Network Voice CAC | p. 294 |
4.4 The Integrated Services Architecture/RSVP | p. 303 |
4.4.1 RSVP | p. 304 |
4.4.2 RSVP Example Reservation Setup | p. 307 |
4.4.3 Application Signaling Interaction | p. 314 |
4.4.4 Intserv over Diffserv | p. 316 |
4.4.5 RSVP Aggregation | p. 320 |
4.4.6 RSVP Traffic Engineering | p. 325 |
4.5 NSIS | p. 326 |
4.6 End-system Measurement-based Admission Control | p. 328 |
4.7 Summary | p. 329 |
References | p. 330 |
5 SLA and Network Monitoring | p. 335 |
5.1 Introduction | p. 335 |
5.2 Passive Network Monitoring | p. 336 |
5.2.1 How Often to Poll? | p. 337 |
5.2.2 Per-link Statistics | p. 337 |
5.2.2.1 Monitoring Classification | p. 338 |
5.2.2.2 Monitoring Policing | p. 339 |
5.2.2.3 Monitoring Queuing and Dropping | p. 342 |
5.2.3 System Monitoring | p. 346 |
5.2.4 Core Traffic Matrix | p. 347 |
5.3 Active Network Monitoring | p. 348 |
5.3.1 Test Stream Parameters | p. 349 |
5.3.1.1 Packet Size | p. 350 |
5.3.1.2 Sampling Strategy | p. 351 |
5.3.1.3 Test Rate | p. 354 |
5.3.1.4 Test Duration and Frequency | p. 355 |
5.3.1.5 Protocols, Ports, and Applications | p. 357 |
5.3.2 Active Measurement Metrics | p. 358 |
5.3.2.1 Delay | p. 358 |
5.3.2.2 Delay-jitter | p. 360 |
5.3.2.3 Packet Loss | p. 362 |
5.3.2.4 Bandwidth and Throughput | p. 363 |
5.3.2.5 Re-ordering | p. 363 |
5.3.2.6 Availability | p. 363 |
5.3.2.7 Quality of Experience | p. 364 |
5.3.3 Deployment Considerations | p. 364 |
5.3.3.1 External versus Embedded Agents | p. 364 |
5.3.3.2 Active Monitoring Topologies | p. 365 |
5.3.3.3 Measuring Equal Cost Multiple Paths | p. 369 |
5.3.3.4 Clock Synchronization | p. 370 |
References | p. 371 |
6 Core Capacity Planning and Traffic Engineering | p. 375 |
6.1 Core Network Capacity Planning | p. 375 |
6.1.1 Capacity Planning Methodology | p. 376 |
6.1.2 Collecting the Traffic Demand Matrices | p. 377 |
6.1.3 Determine Appropriate Over-provisioning Factors | p. 382 |
6.1.4 Simulation and Analysis | p. 388 |
6.2 IP Traffic Engineering | p. 389 |
6.2.1 The Problem | p. 390 |
6.2.2 IGP Metric-based Traffic Engineering | p. 394 |
6.2.3 MPLS Traffic Engineering | p. 397 |
6.2.3.1 MPLS TE Example Tunnel Establishment | p. 397 |
6.2.3.2 Diffserv-aware MPLS Traffic Engineering | p. 404 |
6.2.3.3 MPLS TE Deployment Models and Considerations | p. 408 |
6.2.3.4 Setting Tunnel Bandwidth | p. 412 |
References | p. 414 |
Index | p. 419 |