Telemetry: Industry Status, Challenges, and OSS Opportunities
summary
Telemetry shifts from polling to streaming with model-driven telemetry and YANG models delivered over gNMI/gRPC, enabling real-time automation. OpenConfig and a Telegraf plugin plus Influx/Grafana toolchains are already usable; a TWAMP container demo boots in five minutes. Network metadata via in-situ OAM is implemented in the Linux kernel (more complete than the older VPP port) and silicon proofs exist. Application tracing converges to OpenTelemetry with Jaeger backends, letting trace IDs link app spans to network hops when correlation is exposed.
AI-written summary · about
chapters (9)
- 0:00Intro
- 0:58Network
- 2:59What is telemetry
- 4:20What has changed with telemetry
- 11:46Active monitoring
- 16:56Application level tracing
- 23:31The ultimate idea
- 27:12Open Telemetry
- 28:46Summary
related talks
34mOpenTelemetry: The Vision, Reality, and How to Get Started
50mOpen Telemetry: The Rise of Open Standards in Observability
35mOpenTelemetry for Busy People
20mObservability and Open Telemetry
50mOpen Source Instrumentation - The Rise of OpenTelemetry
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
30mOpenTelemetry 101
18mOpenTelemetry and Observability: What, Why, and Why Now?