Tales of Adopting OpenTelemetry: A First-Timer Journey
summary
Adopts OpenTelemetry to add distributed tracing across Microsoft cloud infra but hits concrete limits: services are C++ and C console apps, communicate via a messaging/batched architecture so context propagation and single-request tracing are nontrivial, automatic instrumentation is limited and much work becomes manual, and some service APIs require modification. Uses the OpenTelemetry Collector and OTLP protocol, starts with a proof of concept, builds shared libraries to reduce dev effort, measures time-to-investigate as the key success metric, and leans on the community for solutions.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:50What is OpenTelemetry
- 8:53Expectations vs reality
- 14:10System landscape overview
- 16:00Integration challenges
- 21:18Adoption hurdles and community
- 27:42Key takeaways
- 29:09Q&A
related talks
37mHow To Adopt OpenTelemetry in an Enterprise Where Incumbent Vendor Tools Reign Supre...
22mBridging the OpenTelemetry Adoption Gap in Cloud-Native Applications
36mObservability For You and Me with OpenTelemetry
01h 03mOpenTelemetry in a Brownfield World
23mModern Application Debugging: An Intro to OpenTelemetry
38mPractical introduction to OpenTelemetry tracing
34mOpenTelemetry: The Vision, Reality, and How to Get Started
01h 08mOpenTelemetry: The First Release, What’s Next, and How to Get Involved