Lightning Talk: Observability-First DevSecOps: Building Resilient Multi-Cloud P...
summary
Shows a concrete pipeline: OpenTelemetry collectors feed Prometheus/Loki and Grafana across AWS, Azure, GCP with Terraform and ArgoCD, and CI runners emit build duration, test failures and deploy latency. Claims MTTR down ~30–35%, cost spikes caught by correlating CPU/memory with billing, and a misconfigured autoscaler in staging fixed before budget impact. Uses team/service labels to route ownership, automates per-environment Grafana dashboards, and enforces golden signals (error rate, latency, saturation) to avoid metric overload.
AI-written summary · about
chapters (10)
- 0:00Introduction to speaker
- 0:37Understanding telemetry
- 2:21Shift to observability-first
- 3:21Architecture and tech stack
- 4:23Embedding telemetry in CI/CD
- 4:47Real-world wins and results
- 6:37Common observability pitfalls
- 8:26Strategic implementation
- 10:03Best practices and lessons
- 12:01Final thoughts and conclusion
related talks
21mtalks about mastering Observability for DevOps
31mCharting New Frontiers: The Evolution of Observability in Cloud Environments
01mProject Lightning Talk: Closing
56mFull-stack observability and application monitoring with
11mLightning Talk: OTel Me How to Build a Data Pipeline for Observability
11mLightning Talk: Beyond Billions: Operating Thanos, Prometheus & OpenTelemetry...
10mLightning Talk: Observability Debt: When Telemetry Stops Telling the Truth
10m⚡ Lightning Talk: Unlocking Customer-Centric Observability: A Case Study of OpenTelemetry To Reduce MTTD < 3 Mins