Book Overflow OTel at Scale - Mastering OpenTelemetry and Observatibilty by Steve Flanders
Oct 6, 2025
Carter and Nathan dive into the intricacies of OpenTelemetry and observability. They discuss cost frameworks, the challenges of vendor billing surprises, and the pros and cons of self-hosting versus managed solutions. The hosts explore observability maturity levels, from basic monitoring to nearly autonomous systems, and highlight the importance of consistent naming conventions to enhance query efficiency. They also touch on future trends influenced by AI and automation, and provide practical insights for engineers looking to master observability.
AI Snips
Chapters
Books
Transcript
Episode notes
Choose Build, Buy, Or Manage Deliberately
- Decide early whether to buy, build, or self-manage your observability stack and budget accordingly.
- Prefer buying unless you have scale or product reasons to self-host, because ops costs and watch-the-watcher risks are high.
Treat Observability As Planned Spend
- Observability spend should be planned as a percentage of cloud or COGS to avoid shock billing.
- Proper budgeting unlocks maturity stages and prevents vendors from blocking advanced features.
Model Observability Costs Proactively
- Measure costs with calculators and add buffers (25–50%) when planning observability spend.
- Use cloud vendor calculators or build cost models to predict and control monthly bills.
