Оглавление
Alex Marcham. Understanding Infrastructure Edge Computing
Table of Contents
List of Tables
List of Illustrations
Guide
Pages
Understanding Infrastructure Edge Computing. Concepts, Technologies and Considerations
Preface. How to Use This Book
About This Book
Audience
About the Author
Acknowledgements
1 Introduction
2 What Is Edge Computing? 2.1 Overview
2.2 Defining the Terminology
2.3 Where Is the Edge?
2.3.1 A Tale of Many Edges
2.3.2 Infrastructure Edge
2.3.3 Device Edge
2.4 A Brief History
2.4.1 Third Act of the Internet
2.4.1.1 The First Act of the Internet
2.4.1.2 The Second Act of the Internet
2.4.1.3 The Third Act of the Internet
2.4.2 Network Regionalisation
2.4.3 CDNs and Early Examples
2.5 Why Edge Computing?
2.5.1 Latency
2.5.2 Data Gravity
2.5.3 Data Velocity
2.5.4 Transport Cost
2.5.5 Locality
2.6 Basic Edge Computing Operation
2.7 Summary
References
3 Introduction to Network Technology. 3.1 Overview
3.2 Structure of the Internet
3.2.1 1970s
3.2.2 1990s
3.2.3 2010s
3.2.4 2020s
3.2.5 Change over Time
3.3 The OSI Model
3.3.1 Layer 1
3.3.2 Layer 2
3.3.3 Layer 3
3.3.4 Layer 4
3.3.5 Layers 5, 6, and 7
3.4 Ethernet
3.5 IPv4 and IPv6
3.6 Routing and Switching
3.6.1 Routing
3.6.2 Routing Protocols
3.6.3 Routing Process
3.6.3.1 Switching
3.6.3.2 Network Boundaries
3.7 LAN, MAN, and WAN
3.8 Interconnection and Exchange
3.9 Fronthaul, Backhaul, and Midhaul
3.10 Last Mile or Access Networks
3.11 Network Transport and Transit
3.12 Serve Transit Fail (STF) Metric
3.13 Summary
References
4 Introduction to Data Centre Technology. 4.1 Overview
4.2 Physical Size and Design
4.3 Cooling and Power Efficiency
4.4 Airflow Design
4.5 Power Distribution
4.6 Redundancy and Resiliency
4.7 Environmental Control
4.8 Data Centre Network Design
4.9 Information Technology (IT) Equipment Capacity
4.10 Data Centre Operation
4.10.1 Notification
4.10.2 Security
4.10.3 Equipment Deployment
4.10.4 Service Offerings
4.10.5 Managed Colocation
4.11 Data Centre Deployment
4.11.1 Deployment Costing
4.11.2 Brownfield and Greenfield Sites
4.11.3 Other Factors
4.12 Summary
References
5 Infrastructure Edge Computing Networks. 5.1 Overview
5.2 Network Connectivity and Coverage Area
5.3 Network Topology
5.3.1 Full Mesh
5.3.2 Partial Mesh
5.3.3 Hub and Spoke
5.3.4 Ring
5.3.5 Tree
5.3.6 Optimal Topology
5.3.7 Inter‐area Connectivity
5.4 Transmission Medium
5.4.1 Fibre
5.4.2 Copper
5.4.3 Wireless
5.5 Scaling and Tiered Network Architecture
5.6 Other Considerations
5.7 Summary
6 Infrastructure Edge Data Centres. 6.1 Overview
6.2 Physical Size and Design
6.2.1 Defining an Infrastructure Edge Data Centre
6.2.2 Size Categories
6.2.2.1 Size Category 1 (<1 kW)
6.2.2.2 Size Category 2 (1–10 kW)
6.2.2.3 Size Category 3 (10–50 kW)
6.2.2.4 Size Category 4 (50–100 kW)
6.2.2.5 Size Category 5 (100–200 kW)
6.2.2.6 Size Category 6 (200–250 kW)
6.2.2.7 Size Category Interoperation
6.3 Heating and Cooling
6.4 Airflow Design
6.4.1 Traditional Designs
6.4.2 Non‐traditional Designs
6.5 Power Distribution
6.6 Redundancy and Resiliency
6.6.1 Electrical Power Delivery and Generation
6.6.2 Network Connectivity
6.6.3 Cooling Systems
6.6.4 Market Design
6.6.5 Redundancy Certification
6.6.6 Software Service Resiliency
6.6.7 Physical Redundancy
6.6.8 System Resiliency Example
6.7 Environmental Control
6.8 Data Centre Network Design
6.9 Information Technology (IT) Equipment Capacity
6.9.1 Operational Headroom
6.10 Data Centre Operation
6.10.1 Site Automation
6.10.2 Single or Multi‐tenant
6.10.3 Neutral Host
6.10.4 Network Operations Centre (NOC)
6.11 Brownfield and Greenfield Sites
6.12 Summary
7 Interconnection and Edge Exchange. 7.1 Overview
7.2 Access or Last Mile Network Interconnection
7.3 Backhaul and Midhaul Network Interconnection
7.4 Internet Exchange
7.5 Edge Exchange
7.6 Interconnection Network Technology
7.6.1 5G Networks
7.6.2 4G Networks
7.6.3 Cable Networks
7.6.4 Fibre Networks
7.6.5 Other Networks
7.6.6 Meet Me Room (MMR)
7.6.7 Cross Connection
7.6.8 Virtual Cross Connection
7.6.9 Interconnection as a Resource
7.7 Peering
7.8 Cloud On‐ramps
7.9 Beneficial Impact
7.9.1 Latency
7.9.2 Data Transport Cost
7.9.3 Platform Benefit
7.10 Alternatives to Interconnection
7.11 Business Arrangements
7.12 Summary
8 Infrastructure Edge Computing Deployment. 8.1 Overview
8.2 Physical Facilities
8.3 Site Locations
8.3.1 kW per kM2
8.3.2 Customer Facility Selection
8.3.3 Site Characteristics
8.4 Coverage Areas
8.5 Points of Interest
8.6 Codes and Regulations
8.7 Summary
9 Computing Systems at the Infrastructure Edge. 9.1 Overview
9.2 What Is Suitable?
9.3 Equipment Hardening
9.4 Rack Densification
9.4.1 Heterogenous Servers
9.4.2 Processor Densification
9.4.3 Supporting Equipment
9.5 Parallel Accelerators
9.5.1 Field Programmable Gate Arrays (FPGAs)
9.5.2 Tensor Processing Units (TPUs)
9.5.3 Graphics Processing Units (GPUs)
9.5.4 Smart Network Interface Cards (NICs)
9.5.5 Cryptographic Accelerators
9.5.6 Other Accelerators
9.5.7 FPGA, TPU, or GPU?
9.6 Ideal Infrastructure
9.6.1 Network Compute Utilisation
9.7 Adapting Legacy Infrastructure
9.8 Summary
References
10 Multi‐tier Device, Data Centre, and Network Resources. 10.1 Overview
10.2 Multi‐tier Resources
10.3 Multi‐tier Applications
10.4 Core to Edge Applications
10.5 Edge to Core Applications
10.6 Infrastructure Edge and Device Edge Interoperation
10.7 Summary
11 Distributed Application Workload Operation. 11.1 Overview
11.2 Microservices
11.3 Redundancy and Resiliency
11.4 Multi‐site Operation
11.5 Workload Orchestration
11.5.1 Processing Requirements
11.5.2 Data Storage Requirements
11.5.3 Network Performance Requirements
11.5.4 Application Workload Cost Profile
11.5.5 Redundancy and Resiliency Requirements
11.5.6 Resource Marketplaces
11.5.7 Workload Requirement Declaration
11.6 Infrastructure Visibility
11.7 Summary
12 Infrastructure and Application Security. 12.1 Overview
12.2 Threat Modelling
12.3 Physical Security
12.4 Logical Security
12.5 Common Security Issues
12.5.1 Staff
12.5.2 Visitors
12.5.3 Network Attacks
12.6 Application Security
12.7 Security Policy
12.8 Summary
13 Related Technologies. 13.1 Overview
13.2 Multi‐access Edge Computing (MEC)
13.3 Internet of Things (IoT) and Industrial Internet of Things (IIoT)
13.4 Fog and Mist Computing
13.5 Summary
Reference
14 Use Case Example : 5G. 14.1 Overview
14.2 What Is 5G?
14.2.1 5G New Radio (NR)
14.2.1.1 Remote Radio Unit (RRU)
14.2.1.2 Distributed Unit (DU)
14.2.1.3 Centralised Unit (CU)
14.2.1.4 Functional Splits
14.2.2 5G Core Network (CN)
14.2.2.1 User Plane Function (UPF)
14.2.2.2 Control Plane
14.3 5G at the Infrastructure Edge
14.3.1 Benefits
14.3.2 Architecture
14.3.3 Considerations
14.4 Summary
15 Use Case Example : Distributed AI. 15.1 Overview
15.2 What Is AI?
15.2.1 Machine Learning (ML)
15.2.2 Deep Learning (DL)
15.3 AI at the Infrastructure Edge
15.3.1 Benefits
15.3.2 Architecture
15.3.3 Considerations
15.4 Summary
16 Use Case Example : Cyber‐physical Systems. 16.1 Overview
16.2 What Are Cyber‐physical Systems?
16.2.1 Autonomous Vehicles
16.2.2 Drones
16.2.3 Robotics
16.2.4 Other Use Cases
16.3 Cyber‐physical Systems at the Infrastructure Edge. 16.3.1 Benefits
16.3.2 Architecture
16.3.3 Considerations
16.4 Summary
Reference
17 Use Case Example : Public or Private Cloud. 17.1 Overview
17.2 What Is Cloud Computing?
17.2.1 Public Clouds
17.2.2 Private Clouds
17.2.3 Hybrid Clouds
17.2.4 Edge Cloud
17.3 Cloud Computing at the Infrastructure Edge. 17.3.1 Benefits
17.3.2 Architecture
17.3.3 Considerations
17.4 Summary
18 Other Infrastructure Edge Computing Use Cases. 18.1 Overview
18.2 Near Premises Services
18.3 Video Surveillance
18.4 SD‐WAN
18.5 Security Services
18.6 Video Conferencing
18.7 Content Delivery
18.8 Other Use Cases
18.9 Summary
19 End to End : An Infrastructure Edge Project Example. 19.1 Overview
19.2 Defining Requirements
19.2.1 Deciding on a Use Case
19.2.2 Determining Deployment Locations
19.2.3 Identifying Required Equipment
19.2.4 Choosing an Infrastructure Edge Computing Network Operator
19.2.5 Regional or National Data Centres
19.3 Success Criteria
19.4 Comparing Costs
19.5 Alternative Options
19.6 Initial Deployment
19.7 Ongoing Operation
19.7.1 SLA Breaches
19.8 Project Conclusion
19.9 Summary
20 The Future of Infrastructure Edge Computing. 20.1 Overview
20.2 Today and Tomorrow
20.3 The Next Five Years
20.4 The Next 10 Years
20.5 Summary
21 Conclusion
Appendix A Acronyms and Abbreviations
Index. a
b
c
d
e
f
g
i
j
l
m
n
o
p
q
r
s
t
u
v
w
x
WILEY END USER LICENSE AGREEMENT
