Founder/CTO Perspectives: The Future of Distributed Tracing
summary
Compares Honeycomb, Sentry and Stanza approaches and makes concrete claims: arbitrarily wide structured data becomes the canonical source of truth from which metrics are derived, auto-instrumentation is the biggest practical gap (especially JavaScript/mobile), and OpenTelemetry is useful but immature for clients. Describes a working UI feature that bubble-up diffs dimensions to find root causes, experiments with injecting continuous profiling to fill span gaps, and recommends incremental rollout—instrument as you debug rather than big rip‑and‑replace.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 4:03Distributed tracing purpose
- 12:52Debugging with observability
- 15:41OpenTelemetry challenges
- 23:08Logs insufficient for debugging
- 25:34eBPF in observability
- 28:38Client-side instrumentation
- 31:04APM tools critique
- 34:39Switch to distributed tracing
- 37:54AI Ops future
- 39:39Tracing versus observability
- 41:05Closing & Q&A
related talks
01h 28mUnderstanding Distributed Systems with OpenTelemetry
44mObservability: Superpowers for Developers
42mObservable Applications Using OpenTelemetry
45mMastering Distributed Tracing: Insights from Building a Platform from Scratch
37mObservability, Distributed Tracing & the Complex World
31mObservability for Developers: How to Get There
57mYour Distributed System is Talking to You: OpenTelemetry
41mModern Application Debugging: An Introduction To OpenTelemetry
