Introducing a Lightweight Rust OpenTelemetry Collector
summary
Rotel is a Rust OpenTelemetry collector that ships as a tiny, open source binary and targets serverless, edge, and high‑scale sidecar use cases. It accepts and exports OTLP natively, supports ClickHouse and Datadog/X‑Ray exporters, and offers an AWS Lambda layer with 60 ms cold start and adaptive flushing. Packaged runtimes (pip/npm) fork the bundled binary and a Python processor API via pyo3 runs user processors with about 75% less memory and ~50% CPU versus the Go collector.
AI-written summary · about
chapters (12)
- 0:00Introducing Rotel and team
- 1:05OpenTelemetry adoption goals
- 2:30Why lightweight collectors
- 5:19Rotel key features
- 8:11Performance benchmarks Lambda
- 9:59Rotel Lambda layer flushing
- 12:05Developer experience embedded
- 13:51Python processor API FFI
- 18:47Python processor performance
- 20:12Getting started Rotel
- 20:33Future roadmap
- 22:46Q&A
related talks
13mManaging Observability Data at the Edge with the OpenTelemetry Collector and OTTL
05mProject Lightning Talk: New Frontiers for OpenTelemetry - A Roadmap for the Future
23mObservability Without Borders: The OpenTelemetry Collector in a WebA...
18mOpenTelemetry and Observability: What, Why, and Why Now?
30mOpenTelemetry 101
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
27mTop 5 Things to Know Running OpenTelemetry on Kubernetes
01h 08mOpenTelemetry: The First Release, What’s Next, and How to Get Involved