OpenTelemetry

Vendor-neutral observability: instrument services once for traces, metrics and logs, propagate context across services, and process telemetry with the OpenTelemetry Collector before sending it to any backend.

Start course →

What you'll learn

  • Explain OpenTelemetry's parts: APIs, SDKs, the Collector, OTLP and semantic conventions.
  • Describe traces, spans and context propagation, and read a distributed trace.
  • Instrument services with zero-code agents and manual spans, metrics and correlated logs.
  • Build OpenTelemetry Collector pipelines with receivers, processors and exporters.
  • Choose sampling strategies and protect data quality, cost and privacy in production.

Syllabus

OpenTelemetry Foundations

  1. Why OpenTelemetry Exists
  2. Signals: Traces, Metrics and Logs
  3. Architecture: API, SDK, Collector, OTLP

Distributed Tracing Concepts

  1. Traces and Spans
  2. Attributes, Events and Status
  3. Context Propagation

Instrumenting Applications

  1. Zero-Code Instrumentation
  2. Manual Spans and Attributes
  3. Resources and Semantic Conventions

Metrics and Logs With OpenTelemetry

  1. Metric Instruments
  2. Logs and Trace Correlation
  3. Linking Metrics to Traces

The OpenTelemetry Collector

  1. Receivers, Processors, Exporters
  2. Collector Configuration
  3. Deployment Patterns

Sampling and Data Processing

  1. Head Sampling
  2. Tail Sampling
  3. Enrichment, Filtering and Redaction

Backends and Debugging With Traces

  1. Choosing Backends
  2. Debugging Latency and Errors With Traces
  3. OpenTelemetry on Kubernetes

OpenTelemetry in Production

  1. Overhead, Volume and Cost
  2. Security and Privacy
  3. Rolling Out OpenTelemetry
  4. An OpenTelemetry Checklist