Decoding OpenTelemetry(OTel) Collector: A Deep Dive into Architecture
summary
Explains how the Collector builds YAML-defined pipelines into a graph and starts them in reverse topological order so exporters start before receivers, and shuts down in the opposite order to avoid data loss. Describes building custom distributions with OCB manifest that mixes upstream and private components, details the exporterhelper defaults (in-memory queue, 5s initial retry, 300s max retry) and how to switch to a persistent queue via a file storage extension, and contrasts exporterhelper with the batch processor.
AI-written summary · about
chapters (11)
- 0:00OpenTelemetry Collector intro
- 1:07OTel Collector basics
- 2:33Core Collector components
- 5:47Collector component myths
- 8:40Custom Collector builds
- 10:27Data pipeline flow
- 11:54Internal telemetry
- 14:03Exporter helper library
- 17:40Helper vs batch processor
- 19:55Community updates
- 21:20Question and Answer Session
related talks
25mDebugging the Pipeline: Observability for Your Observability with OTel + Prometheus
01h 03mGet a Grip on your Telemetry using the OpenTelemetry Collector
13mManaging Observability Data at the Edge with the OpenTelemetry Collector and OTTL
26mSimplifying OpenTelemetry with Configuration
28mControlling data overhead from the OpenTelmetry Collector
31mOne Binary, Two Ecosystems: Embedding Prometheus Exporters With OCB
11mLightning Talk: OTel Me How to Build a Data Pipeline for Observability
09mOpenTelemetry or OpenTeleMessy?