From Flaky Builds to Self-Healing Deployments: An OTel-Driven CI/CD...
summary
Claims a production CI/CD redesign that links build metadata to runtime traces across 60M+ RPM clusters and turns observability into actionable fixes. Uses GitHub Actions/Jenkins for CI, Argo Rollouts for delivery, OpenTelemetry with Jaeger/Prometheus/Loki for tracing and metrics, and Grafana MCP plus a custom AI SRE agent to fetch traces, open PRs and trigger GitOps rollbacks. Injected build IDs into pods, added Argo analysis templates querying OTEL, demoed an induced 0.3 error rate and automated a fix that reduces failures to zero.
AI-written summary · about
chapters (13)
- 0:00Flaky builds and self-healing
- 1:03Traditional CI/CD problems
- 4:33Demo of real-world issue
- 7:40Trace IDs for debugging
- 8:01OTel-driven CI/CD re-architecture
- 9:51Self-healing with AI SRE agents
- 11:16Distributed tracing for fixes
- 13:06OTel impact on debugging
- 15:36Live AI SRE agent demo
- 25:35Key benefits and ecosystem
- 26:59Takeaways and best practices
- 27:41Flaky test detection extension
- 29:05Closing and Q&A
related talks
28mEnd-to-end CI/CD observability with open source
22mThe Ortelius Journey with GitOps and Event Driven CI/CD
51mSelf-Healing CI/CD Pipelines with Tekton and ArgoCD
25mBeyond Production: End-to-End CI/CD Observability With OpenTelemetry
44mAgentic Delivery, Guardrailed: AI in CI/CD & Platform Engineering (Open Source Patterns)
06mLightning Talk: Not Just Enterprise. Modern Java App CI/CD Observability with OTel...
13mLightning Talk: Observability-First DevSecOps: Building Resilient Multi-Cloud P...
30mOur Journey in Integrating GitOps Tools Into Delivery Pipelines