Practical OpenTelemetry with Prometheus 3.0
summary
Prometheus 3.0 now accepts OpenTelemetry metrics via the stable OTLP endpoint and recommends enabling out_of_order mode because collectors batch and replicas cause interleaving; a new info() function joins job and instance to pull resource attributes at query time (MVP under review for 3.3.x) so you avoid series churn from promoting labels; a config section lets you promote a curated list of ~17 attributes on ingestion; translation code is optimized and is reported 40% faster and 60% less memory.
AI-written summary · about
chapters (7)
- 0:00Introduction and setup
- 1:07Resource attributes handling
- 9:47UTF-8 support progress
- 12:13OTLP input stability
- 15:49Delta temporality explained
- 19:00Future plans and guide
- 20:07Q&A session
related talks
28mOpenTelemetry developments in Prometheus over the last year
05mLightning Talk: Practical OpenTelemetry with Prometheus 3.0
27mWelcome + Project Updates
30mOpenTelemetry 101
06mLightning Talk: OTel Demo: Showcasing Prometheus and improving its OTel support
36mObservability For You and Me with OpenTelemetry
18mMaking Prometheus even more open
19mOpenTelemetry's Path to Observability's Future