Practical OpenTelemetry

Learn the value that OpenTelemetry can bring to organizations that aim to implement observability best practices, and gain a deeper understanding of how different building blocks interact with each other to bring out-of-the-box, vendor-neutral instrumentation to your stack. With examples in Java, this book shows how to use OpenTelemetry APIs and configure plugins and SDKs to instrument services and produce valuable telemetry data. You’ll learn how to maximize adoption of OpenTelemetry and encourage the change needed in debugging workflows to reduce cognitive load for engineers troubleshooting production workloads.

Adopting observability best practices across an organization is challenging. This book begins with a discussion of how operational monitoring processes widely followed for decades fall short at providing the insights needed for debugging cloud-native, distributed systems in production. The book goes on to show how the Cloud Native Computing Foundation’s OpenTelemetry project helps you standardize instrumentation and transport of telemetry signals, providing a common language for all observability tooling.

YOU WILL LEARN

* Why observability is a necessity in modern distributed systems
* The value of OpenTelemetry for engineers and organizations
* OpenTelemetry component specification and general design
* Tracing, metrics, and logs APIs and SDKs, with examples in Java
* OpenTelemetry Collectors and recommended transport and processing pipelines
* How to adopt observability standards across an organization

WHO THIS BOOK IS FOR

Software engineers familiar with cloud-native technologies and operational monitoring who want to instrument and export telemetry data from their services; observability leads who want to roll out OpenTelemetry standards and best practices across their organizations; and Java developers who want a book with OpenTelemetry examples in that language

DANIEL GOMEZ BLANCO is a Principal Engineer at Skyscanner, leading their observability transformation across hundreds of services to ensure that travelers get a reliable and performant experience when booking their next holiday. He is an advocate of open standards and CNCF projects such as OpenTelemetry to back the instrumentation and collection of operational data. Daniel has experience working in organizations of all sizes, from international institutions such as CERN in Geneva, to London startups such as SKIPJAQ. His main focus has always been building software and adopting solutions to minimize the cognitive load required for engineers to support and operate production services. PART I. THE NEED FOR OBSERVABILITY WITH OPENTELEMETRY

1. The Need for Observability

a. Why Observability Matters

b. Context and Correlation

2. How OpenTelemetry Enables Observability

a. OpenTelemetry’s Mission

b. The Power of Open Standards

c. The Shift In Vendor Added Value

PART II. OPENTELEMETRY COMPONENTS AND BEST PRACTICES

3. OpenTelemetry Fundamentals

a. OpenTelemetry Specification

b. Semantic Conventions

4. Auto-Instrumentation

a. Resource SDK

b. Instrumentation Libraries

5. Context, Baggage, and Propagators

a. Telemetry Context and the Context API

b. Baggage API

c. Cross-Service Context and the Propagators API

6. Tracing

a. What is a Distributed Trace?

b. Tracing API

c. Tracing SDK

d. Trace Context Propagation

7. Metrics

a. Measurements, Metrics and Time Series

b. Metrics API

c. Metrics SDK

8. Logging

a. The Purpose of Logs for Observability

b. Logging API

c. Logging SDK

d. Integrations with Logging Frameworks

9. Protocol and Collector

a. Protocol

b. Collector

10. Sampling and Common Deployment Models

a. Common Deployment Models

b. Trace Sampling

PART III. ROLLING OUT OPENTELEMETRY ACROSS YOUR ORGANISATION

11. Maximizing Adoption by Minimizing Friction

a. Investing in Telemetry Enablement

b. Adopting OpenTelemetry

12. Adopting Observability

a. Shifting Debugging Workflows

b. Expanding Context

c. Keeping Telemetry Valuable