Book Overflow OTel at Scale - Mastering OpenTelemetry and Observatibilty by Steve Flanders
13 snips
Oct 20, 2025 Carter and Nathan dive into the intricacies of OpenTelemetry and observability, highlighting Steve Flanders' insights. They explore the importance of choosing the right observability platform and discuss real-world cost surprises with vendors. The conversation evolves into the five maturity levels of observability, from basic monitoring to autonomous systems, illustrating the implications of each stage. They also tackle common pitfalls like vendor lock-in and alert noise, while speculating on future trends shaped by AI and evolving compliance needs.
AI Snips
Chapters
Books
Transcript
Episode notes
Budget Observability Like A Product
- Establish a clear budget for observability and treat it like a real business expense.
- Track expected spend (e.g., 1–15% of cloud/COGS) and set a sensible buffer to avoid surprises.
Self-Hosting Adds Hidden Ops Costs
- Self-hosting observability shifts significant operational burdens onto your team.
- That creates an added dependency: who monitors the observability stack when it fails?
Estimate Vendor Costs Conservatively
- Use vendor cost calculators and include buffers (25–50%) when estimating observability bills.
- Monitor usage monthly and refine estimates to avoid painful surprises.
