Practical Observability Engineering
Better signal, not more telemetry.
Olatuak is an independent technical brand focused on helping engineers build, break, debug, and scale real-world observability systems using OpenTelemetry.
I build, break, debug, and scale observability systems. With hands-on experience spanning OpenTelemetry, Kubernetes, telemetry pipelines, and earlier deep work with Datadog, I write and teach from direct engineering experience.
Most teams do not struggle because they lack data. They struggle because telemetry is noisy, expensive, poorly governed, and hard to turn into confident operational decisions. Olatuak exists to cut through that noise.
The Three Pillars of Olatuak
How to implement observability systems properly
Collector topology, pipelines, semantic conventions, instrumentation, and Kubernetes architectures.
Why observability systems fail and how to diagnose them
Spans disappearing, broken context propagation, backpressure, dropped telemetry, and queue bottlenecks.
Observability across teams, fleets, and high-volume systems
Telemetry governance, fleet management, cardinality controls, sampling at scale, and cost efficiency.
Editorial Philosophy
Olatuak follows a simple rule: take a position, explain why, and show the exceptions. You won't find generic feature summaries or documentation copy-pastes here. Instead, expect:
- Real engineering problems: reproducing difficult edge cases and subtle bugs.
- Isolated experiments: stress-testing Collector queues, batching, and backpressure.
- Architecture & governance: naming standards, ownership metadata, and sampling strategies.
- Cost containment: identifying wasted telemetry before it turns into an invoice shock.
Where to Follow
The Olatuak engine spans three channels:
- YouTube (Discovery & Experiments): Demonstrations, troubleshooting walkthroughs, and live architectural experiments.
- The Blog (Depth & Reference): Field guides, configuration patterns, commands, diagrams, and exact technical references.
- GitHub (Reproducibility & Labs): Ready-to-run configurations and docker-compose labs to reproduce issues locally.