Implementing Unified Observability: IDP on Backstage wit...
summary
Shows how Backstage backend (Node.js) is instrumentation.js with node auto-instrumentation to emit traces and metrics to an OpenTelemetry Collector → Signals pipeline, with trace IDs injected into logs for correlation. Demonstrates a custom UUID generator plugin, a GitHub Actions → Argo CD deployment flow, alerting (Slack/email), and troubleshooting: Argo CD API calls take ~3380 ms and produce 49 spans, and observed 400/500 errors during failed deploys that trigger alerts.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 1:02Monitor Backstage with OpenTelemetry
- 3:27Backstage architecture overview
- 5:31OpenTelemetry and ecosystem
- 8:44Unified telemetry benefits
- 9:05Scaling Backstage monitoring
- 10:30Debugging with traces and metrics
- 12:39Live demo setup
- 14:01Instrumenting Backstage backend
- 15:28Custom plugin integration
- 17:17Backstage backend logs
- 18:42New API component deployment
- 19:26Analyzing traces and alerts
- 20:53Debugging server errors
- 21:35Performance bottlenecks analysis
- 22:19Summary
related talks
26mObservability for Your Internal Developer Platform: Instrumenting Bac...
15mPlatforming the Platform: Running Our IDP With Our IDP
40mObservability for All!
53mUnmask hidden issues with Observability superpowers!
32mObservability for Frontend Engineers
51mFull-Stack Observability: One Tool to Rule Them All
17mUnlocking Observability with React & Node.js
31mObservability for Developers: How to Get There