| |
| |
Preface | |
| |
| |
Acknowledgments | |
| |
| |
About the authors | |
| |
| |
| |
QOS Requirements and Service Level Agreements | |
| |
| |
| |
Introduction | |
| |
| |
| |
SLA Metrics | |
| |
| |
| |
Network Delay | |
| |
| |
| |
Propagation Delay | |
| |
| |
| |
Switching Delay | |
| |
| |
| |
Scheduling Delay | |
| |
| |
| |
Serialization Delay | |
| |
| |
| |
Delay-jitter | |
| |
| |
| |
Packet Loss | |
| |
| |
| |
Bandwidth and Throughput | |
| |
| |
| |
Layer 2 Overheads | |
| |
| |
| |
VPN Hose and Pipe Models | |
| |
| |
| |
Per Flow Sequence Preservation | |
| |
| |
| |
Availability | |
| |
| |
| |
Network Availability | |
| |
| |
| |
Service Availability | |
| |
| |
| |
Quality of Experience | |
| |
| |
| |
Voice | |
| |
| |
| |
Video | |
| |
| |
| |
On-line Gaming | |
| |
| |
| |
Application SLA Requirements | |
| |
| |
| |
Voice over IP | |
| |
| |
| |
VoIP: Impact of Delay | |
| |
| |
| |
VoIP: Impact of Delay-jitter | |
| |
| |
| |
VoIP: Impact of Loss | |
| |
| |
| |
VoIP: Impact of Throughput | |
| |
| |
| |
VoIP: Impact of Packet Re-ordering | |
| |
| |
| |
Video | |
| |
| |
| |
Video Streaming | |
| |
| |
| |
Video Conferencing | |
| |
| |
| |
Data Applications | |
| |
| |
| |
Throughput Focussed TCP Applications | |
| |
| |
| |
Interactive Data Applications | |
| |
| |
| |
On-line Gaming | |
| |
| |
| |
Marketed SLAs versus Engineered SLAs | |
| |
| |
| |
End-to-End SLAs vs Segmented SLAs | |
| |
| |
| |
Inter-provider SLAs | |
| |
| |
| |
Intserv and Diffserv SLAs | |
| |
| |
References | |
| |
| |
| |
Introduction to QOS Mechanics and Architectures | |
| |
| |
| |
What is Quality of Service? | |
| |
| |
| |
Quality of Service vs Class of Service or Type of Service? | |
| |
| |
| |
Best-effort Service | |
| |
| |
| |
The Timeframes that Matter for QOS | |
| |
| |
| |
Why IP QOS? | |
| |
| |
| |
The QOS Toolset | |
| |
| |
| |
Data Plane QOS Mechanisms | |
| |
| |
| |
Classification | |
| |
| |
| |
Implicit Classification | |
| |
| |
| |
Complex Classification | |
| |
| |
| |
Deep Packet Inspection/Stateful Inspection | |
| |
| |
| |
Simple Classification | |
| |
| |
| |
Marking | |
| |
| |
| |
Policing and Metering | |
| |
| |
| |
RFC 2697: Single Rate Three Color Marker | |
| |
| |
| |
RFC 2698: Two Rate Three Color Marker | |
| |
| |
| |
Color-aware Policers | |
| |
| |
| |
Metering | |
| |
| |
| |
Queuing, Scheduling, Shaping, and Dropping | |
| |
| |
| |
Queuing and Scheduling | |
| |
| |
| |
Dropping | |
| |
| |
| |
Shaping | |
| |
| |
| |
Link Fragmentation and Interleaving | |
| |
| |
| |
IP QOS Architectures | |
| |
| |
| |
A Short History of IP Quality of Service | |
| |
| |
| |
Type of Service/IP Precedence | |
| |
| |
| |
IP Precedence | |
| |
| |
| |
Type of Service | |
| |
| |
| |
IPv6 Traffic Class Octet | |
| |
| |
| |
Integrated Services Architecture | |
| |
| |
| |
Differentiated Services Architecture | |
| |
| |
| |
DS Field | |
| |
| |
| |
Per-Hop Behaviors | |
| |
| |
| |
Per-Domain Behaviors | |
| |
| |
| |
Explicit Congestion Notification | |
| |
| |
| |
Diffserv Tunneling Models | |
| |
| |
| |
IPv6 QOS Architectures | |
| |
| |
| |
MPLS QOS Architectures | |
| |
| |
| |
MPLS and Intserv/RSVP | |
| |
| |
| |
MPLS and Diffserv | |
| |
| |
| |
IP Multicast and QOS | |
| |
| |
| |
Typical Router QOS Implementations in Practice | |
| |
| |
| |
Layer 2 QOS | |
| |
| |
| |
ATM | |
| |
| |
| |
Mapping Diffserv to ATM QOS | |
| |
| |
| |
Frame-relay | |
| |
| |
| |
Ethernet | |
| |
| |
| |
Complementary Technologies | |
| |
| |
| |
Where QOS cannot make a difference | |
| |
| |
References | |
| |
| |
| |
Precedence, TOS, and DSCP Conversion | |
| |
| |
| |
Notation | |
| |
| |
| |
Conversion | |
| |
| |
| |
Deploying Diffserv | |
| |
| |
| |
Introduction | |
| |
| |
| |
Deploying Diffserv at the Network Edge | |
| |
| |
| |
Why is the Edge Key for Tight SLA Services? | |
| |
| |
| |
Edge Diffserv Case Study | |
| |
| |
| |
SLA Specification | |
| |
| |
| |
Diffserv Meta-Language | |
| |
| |
| |
High-speed Edge Design | |
| |
| |
| |
Design Variations | |
| |
| |
| |
Edge SLA Summary | |
| |
| |
| |
How Many Classes are Enough? | |
| |
| |
| |
What Marking Scheme to Use? | |
| |
| |
| |
VoIP - How Much is Enough at the Edge? | |
| |
| |
| |
Deploying Diffserv in the Network Backbone | |
| |
| |
| |
Is Diffserv Needed in the Backbone? | |
| |
| |
| |
Core Case Study | |
| |
| |
| |
Core Classes of Service and SLA Specification | |
| |
| |
| |
"Prioritized" Diffserv Core Model | |
| |
| |
| |
Detailed Core Design | |
| |
| |
| |
Design Variations | |
| |
| |
| |
Core-marking Scheme | |
| |
| |
| |
Tuning (W)RED | |
| |
| |
| |
Tuning the Exponential Weighting Constant | |
| |
| |
| |
Tuning Minth and Maxth | |
| |
| |
| |
Mark Probability Denominator | |
| |
| |
| |
In- and Out-of-contract | |
| |
| |
References | |
| |
| |
| |
Capacity Admission Control | |
| |
| |
| |
Introduction | |
| |
| |
| |
When is Admission Control Needed? | |
| |
| |
| |
A Taxonomy for Admission Control | |
| |
| |
| |
What Information is Needed for Admission Control? | |
| |
| |
| |
Parameterized or Measurements-based Algorithms | |
| |
| |
| |
Parameterized Algorithms | |
| |
| |
| |
Measurement-based Algorithms | |
| |
| |
| |
Topology-unaware Off-path CAC | |
| |
| |
| |
Topology-aware Off-path CAC: "Bandwidth Manager" | |
| |
| |
| |
Example Bandwidth Manager Method of Operation: Next Generation Network Voice CAC | |
| |
| |
| |
The Integrated Services Architecture/RSVP | |
| |
| |
| |
RSVP | |
| |
| |
| |
RSVP Example Reservation Setup | |
| |
| |
| |
Application Signaling Interaction | |
| |
| |
| |
Intserv over Diffserv | |
| |
| |
| |
RSVP Aggregation | |
| |
| |
| |
RSVP Traffic Engineering | |
| |
| |
| |
NSIS | |
| |
| |
| |
End-system Measurement-based Admission Control | |
| |
| |
| |
Summary | |
| |
| |
References | |
| |
| |
| |
SLA and Network Monitoring | |
| |
| |
| |
Introduction | |
| |
| |
| |
Passive Network Monitoring | |
| |
| |
| |
How Often to Poll? | |
| |
| |
| |
Per-link Statistics | |
| |
| |
| |
Monitoring Classification | |
| |
| |
| |
Monitoring Policing | |
| |
| |
| |
Monitoring Queuing and Dropping | |
| |
| |
| |
System Monitoring | |
| |
| |
| |
Core Traffic Matrix | |
| |
| |
| |
Active Network Monitoring | |
| |
| |
| |
Test Stream Parameters | |
| |
| |
| |
Packet Size | |
| |
| |
| |
Sampling Strategy | |
| |
| |
| |
Test Rate | |
| |
| |
| |
Test Duration and Frequency | |
| |
| |
| |
Protocols, Ports, and Applications | |
| |
| |
| |
Active Measurement Metrics | |
| |
| |
| |
Delay | |
| |
| |
| |
Delay-jitter | |
| |
| |
| |
Packet Loss | |
| |
| |
| |
Bandwidth and Throughput | |
| |
| |
| |
Re-ordering | |
| |
| |
| |
Availability | |
| |
| |
| |
Quality of Experience | |
| |
| |
| |
Deployment Considerations | |
| |
| |
| |
External versus Embedded Agents | |
| |
| |
| |
Active Monitoring Topologies | |
| |
| |
| |
Measuring Equal Cost Multiple Paths | |
| |
| |
| |
Clock Synchronization | |
| |
| |
References | |
| |
| |
| |
Core Capacity Planning and Traffic Engineering | |
| |
| |
| |
Core Network Capacity Planning | |
| |
| |
| |
Capacity Planning Methodology | |
| |
| |
| |
Collecting the Traffic Demand Matrices | |
| |
| |
| |
Determine Appropriate Over-provisioning Factors | |
| |
| |
| |
Simulation and Analysis | |
| |
| |
| |
IP Traffic Engineering | |
| |
| |
| |
The Problem | |
| |
| |
| |
IGP Metric-based Traffic Engineering | |
| |
| |
| |
MPLS Traffic Engineering | |
| |
| |
| |
MPLS TE Example Tunnel Establishment | |
| |
| |
| |
Diffserv-aware MPLS Traffic Engineering | |
| |
| |
| |
MPLS TE Deployment Models and Considerations | |
| |
| |
| |
Setting Tunnel Bandwidth | |
| |
| |
References | |
| |
| |
Index | |