Real-world edge computing : scale, secure, and succeed in the realm of edge computing with Open Horizon / Robert High, Sanjeev Gupta.

Format:

Book

Author/Creator:

High, Robert, author.

Gupta, Sanjeev, author.

Language:

English

Subjects (All):

Edge computing.

Physical Description:

1 online resource (296 pages) : illustrations

Edition:

[First edition].

Place of Publication:

Birmingham, UK : Packt Publishing Ltd., 2024.

Summary:

Edge computing holds vast potential to revolutionize industries, yet its implementation poses unique challenges. Written by industry veterans Rob High and Sanjeev Gupta, this comprehensive guide bridges the gap between theory and practice. Distilling expertise from their combined decades of experience in edge computing and hybrid cloud mesh solutions, this book equips software developers and DevOps teams with the knowledge and skills needed to deploy edge solutions at scale in production environments. It also explores foundational standards and introduces key factors that may impede the scaling of edge solutions. While edge computing draws from the successes of cloud computing, crucial distinctions separate the two. High and Gupta elucidate these distinctions, helping you grasp the nuanced dynamics of edge-computing ecosystems. With a focus on leveraging Open Horizon to overcome pitfalls and optimize performance, this book will help you confidently navigate the intricacies of constructing and deploying resilient edge solutions in real-world production settings. By the end of this book, you’ll have acquired a deep understanding of essential success factors for building and deploying robust edge solutions in real-world production settings, leveraging Open Horizon for scalable edge deployments.

Contents:

Cover

Title Page

Copyright

Dedication

Contributors

Table of Contents

Preface

Part 1: Managing the Edge

Chapter 1: Fundamentals of Edge Computing

Technical requirements

A primer on Edge computing

Why is Edge computing needed?

Applications of Edge computing

Quality inspection

POS

Worker safety

Store automation

Automated teller machines (ATMs)

Facilities inspection and the digital transformation of analog production

Site security and loss prevention

Generalizations - AI is the dominant workload at the Edge

The Edge computing ecosystem

Far Edge - Edge devices

Far Edge - Edge servers

Network and metro Edge

Cloud Edge

Perspectives on other ecosystem participants

Summary

Chapter 2: Practicalities of Edge Computing

Practical challenges with Edge computing

Designing and building Edge software

Application placement and life cycle management

Managing the scale

Autonomous and flexible operations

Security at the edge

Data privacy

Artificial intelligence at the Edge

Connectivity at the Edge

Organization considerations

CloudOps and DevOps roles and responsibilities

Cost-effective architecture

Decision latency

Skills at the Edge of business

Chapter 3: Designing and Building Edge Software

Sample Edge computing infrastructure

General-purpose processing architectures

Specialized processing architectures

The Linux platform

The container runtime

Roles and responsibilities in successful production Edge deployments

Software engineering

AI/ML training (data science)

Infrastructure provisioning

Software and model deployment

The best-practice enterprise architecture for Edge-native solutions

Start with Cloud-native.

The separation of algorithmic logic and ML models

Conditional location independence

Autonomous deployment

Edge connectivity

Chapter 4: Edge Container Platforms

Open Container Initiative (OCI)

Docker

Docker engine

Docker Hub

Podman

Kubernetes

Chapter 5: Application Placement and Life Cycle Management

Deploying containers to the Edge

Introduction to Open Horizon and IEAM

The architecture of the management system

What is a service?

The components of the Management Hub

Agbot

Exchange

Switchboard

Model Manager

Secrets Manager

FDO

Organizations and tenants

More about the Edge Agent

Automatic Agent updates

An examination of service life cycles

An overview of patterns

Policy-based management is key to deployment at scale

Node Policies

Service Policies and Service Definitions

Model Policies

Deployment Policies

Policy-based placement negotiations

Part 2: Working on the Edge

Chapter 6: Installing an Edge Agent on an Edge Node

What is a Management Hub?

What is an Edge Agent?

Supported architectures and operating systems

Options to install the Edge Agent

Automated agent installation

Bulk agent installation

FDO-based agent installation

Preparing the working environment

Setting up a non-root user

Installing AIO Open Horizon Management Hub

Installing the Edge Agent on an Edge device

Chapter 7: Registering an Edge Node

What is Edge node registration?

Verifying the pre-condition

Verifying the current Edge node condition

Registering an Edge device node

Verifying the post-condition

Chapter 8: Using the Open Horizon CLI and API

Technical requirements.

Using the CLI

Commonly used CLIs

General commands

Edge node registration

Service development, publishing, and deployment

Pattern publishing

Deployment policy

Adding a Node policy using the Management Hub

Managing a local Edge Node using Node policy

User input for services

Deployment check

Model management system

APIs

Directly calling APIs

Chapter 9: Autonomous and Flexible Management of Edge Services

Developing an Open Horizon Service

Developing an Open Horizon Pattern

Publishing an Open Horizon Service and Pattern

Deploying a workload on an Edge device node using a pattern

Combining multiple services and deployments

Chapter 10: Managing Edge Scale with Policy-Based Workload Placement

Policy-based application workload placement

Developing a Service policy

Listing a Service policy

Policy-based Properties and Constraints

Creating a Service policy

Adding a Service policy

Viewing the Service policy's details

Adding Service policies for other services

Removing a service policy

Developing a Deployment policy

Creating a Deployment policy

Adding a Deployment policy

Listing a Deployment policy

Adding Deployment policies for other services

Removing a Deployment policy

Developing a Node policy

Creating a Node policy

Applying a Node policy

Modifying the Node policy

Practicing Node policy modifications

Removing policies

Part 3: Advancing the Edge System

Chapter 11: Machine Learning Workload and Model Deployment

Use of MMS

MMS components

CSS

ESS

Model deployment policy metadata

Deploying a Service to consume a published model

Deploying the example ML service

Publishing a model file to an Edge node.

Verifying model file delivery

Chapter 12: Security at the Edge

Security vulnerabilities at the Edge

Masquerading devices

Software tampering

Backdoor attacks

Replay attacks

Unauthorized channels

Hacking

Excessive privilege

Exploiting software vulnerabilities

Operating failure

System inconsistencies

How Open Horizon protects your Edge

Device attestation

Tamper resistance

Intrusion resistance

Replay resistance

Workload isolation

Blast protection

Role-based privileges

Secrets management

Failure resistance

Infrastructure currency

Workload signing

API keys, secrets, and the Vault

Chapter 13: Network Connectivity at the Edge

The shifting enterprise networking paradigm

The history of network management in connected systems

The classic paradigm is inherently contentious

The connected Edge

Connecting the Edge

How does the enterprise network work?

Operating the connected Edge

Deploying the connected Edge

Part 4: Edge Management in Practice

Chapter 14: Building a Real-World Example Application

A typical Edge device node

Machine inferencing pipeline

Services, Policies, and User Input

Services

Policies

User Input

Publishing Services and Deployment policies

Registering the Edge device node with an initial ML model

Registering the Edge device node

Verifying the application

Publishing an ML model using MMS

Chapter 15: Troubleshooting at the Edge

Answers to not-so-common questions

Document references

Edge Agent

Service publishing

Pattern-based service deployment

Docker usage

Uninstalling Open Horizon Management Hub.

Edge device agent verification

Chapter 16: Follow-on Topics

Fleet management

Don't brick your Edgeedge

Air-gapping your Edge

Edge and mobile computing

Building and maintaining a robust connected Edge

What about multi-access Edge Computing?

But wait, what is the network Edge?

Chapter 17: Using the IBM Edge Application Manager Web UI

Logging into the Home screen of IEAM

Navigating the main tabs of the UI

Exploring the Edge nodes in your organization

Node properties

Node constraints

Deployed services

Publishing and editing your services

Creating deployment policies to initiate a deployment of your service

Viewing and managing patterns

Index

About PACKT

Other Books You May Enjoy.

Notes:

Includes index.

OCLC-licensed vendor bibliographic record.

Description based on publisher supplied metadata and other sources.

ISBN:

1-83588-584-5

OCLC:

1444260114