OpenTelemetry is a revolution in observability data. Instead of running multiple uncoordinated pipelines, OpenTelemetry provides users with a single integrated stream of data, providing multiple sources of high-quality telemetry data: tracing, metrics, logs, RUM, eBPF, and more. This practical guide shows you how to set up, operate, and troubleshoot the OpenTelemetry observability system.
Authors Austin Parker, head of developer relations at Lightstep and OpenTelemetry Community Maintainer, and Ted Young, cofounder of the OpenTelemetry project, cover every OpenTelemetry component, as well as observability best practices for many popular cloud, platform, and data services such as Kubernetes and AWS Lambda. You'll learn how OpenTelemetry enables OSS libraries and services to provide their own native instrumentation—a first in the industry.
Ideal for application developers, OSS maintainers, operators and infrastructure teams, and managers and team leaders, this...
Author(s): Ted Young & Austin Parker
Publisher: O'Reilly Media
Year: 2024
Language: English
Pages: 250
Foreword
Preface
Conventions Used in This Book
Using Code Examples
O’Reilly Online Learning
How to Contact Us
Acknowledgments
Austin
Ted
1. The State of Modern Observability
The Times They Are A-Changin’
Observability: Key Terms to Know
A Brief History of Telemetry
The Three Browser Tabs of Observability
Emerging Complications
The Three Pillars Were an Accident
A Single Braid of Data
Conclusion
2. Why Use OpenTelemetry?
Production Monitoring: The Status Quo
The Challenges of Production Debugging
The Importance of Telemetry
Hard and Soft Context
Telemetry Layering
Semantic Telemetry
What Do People Need?
Developers and Operators
Teams and Organizations
Why Use OpenTelemetry?
Universal Standards
Correlated Data
Conclusion
3. OpenTelemetry Overview
Primary Observability Signals
Traces
Metrics
Logs
Observability Context
The Context Layer
Attributes and Resources
Semantic Conventions
OpenTelemetry Protocol
Compatibility and Future-Proofing
Conclusion
4. The OpenTelemetry Architecture
Application Telemetry
Library Instrumentation
The OpenTelemetry API
The OpenTelemetry SDK
Infrastructure Telemetry
Telemetry Pipelines
What’s Not Included in OpenTelemetry
Hands-On with the OpenTelemetry Demo
Running the Demo
Architecture and Design
Managing Application Performance with OpenTelemetry
Finding Needles in Haystacks
Observability Pipelines in the Demo
The New Observability Model
Conclusion
5. Instrumenting Applications
Agents and Automated Setup
Installing the SDK
Registering Providers
Providers
TracerProvider
MeterProvider
LoggerProvider
Shutting Down Providers
Custom Providers
Configuration Best Practices
Remote Configuration
Attaching Resources
Resource Detectors
Service Resources
Installing Instrumentation
Instrumenting Application Code
How Much Is Too Much?
Layering Spans and Metrics
Browser and Mobile Clients
The Complete Setup Checklist
Packaging It All Up
Conclusion
6. Instrumenting Libraries
The Importance of Libraries
Why Provide Native Instrumentation?
Observability Works by Default in Native Instrumentation
Native Instrumentation Lets You Communicate with Your Users
Native Instrumentation Shows That You Care About Performance
Why Aren’t Libraries Already Instrumented?
How OpenTelemetry Is Designed to Support Libraries
OpenTelemetry Separates the Instrumentation API and the Implementation
OTel Maintains Backward Compatibility
OTel Keeps Instrumentation Off by Default
Shared Libraries Checklist
Shared Services Checklist
Conclusion
7. Observing Infrastructure
What Is Infrastructure Observability?
Observing Cloud Providers
Collecting Cloud Metrics and Logs
Metamonitoring
Collectors in Containers
Observing Platforms
Kubernetes Platforms
Serverless Platforms
Queues, Service Buses, and Other Async Workflows
Conclusion
8. Designing Telemetry Pipelines
Common Topologies
No Collector
Local Collector
Collector Pools
Pipeline Operations
Filtering and Sampling
Transforming, Scrubbing, and Versioning
Transforming Telemetry with OTTL
Privacy and Regional Regulations
Buffering and Backpressure
Changing Protocols
Collector Security
Kubernetes
Managing Telemetry Costs
Conclusion
9. Rolling Out Observability
The Three Axes of Observability
Deep Versus Wide
Code Versus Collection
Centralized Versus Decentralized
Moving from Innovation to Differentiation
Observability as Testing
Green Observability
AI Observability
OpenTelemetry Rollout Checklist
Conclusion
A. The OpenTelemetry Project
Organizational Structure
The OpenTelemetry Specification
Project Management
How to Get Involved
Where to Find Us
B. Further Resources
Websites
Books
Index
About the Authors
