Combining Progressive Delivery with GitOps and Continuous Delivery
summary
Shows an end to end GitOps pipeline that builds images with Kaniko, promotes environments by committing to Git rather than touching clusters, and gates canary rollouts via Istio and Prometheus metrics with staged percentages like 5 → 10 → 15 → 20% and an 80% success threshold. Uses Argo CD, Argo Workflows and Argo Events to decouple webhooks from execution, stores secrets with Sealed Secrets, and enforces no human cluster access; demo attempts to exercise the flow but the live demo encounters failures so expect conceptual wiring plus partial running examples.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 7:00Demo
- 9:50Events
- 10:45Even Sources
- 12:05Production Sensors
- 15:55Junk
- 16:40Application
- 16:55Simulation
- 17:35What every developer does
- 18:35What happens in the background
- 19:55I messed up my pipeline
- 27:00Outtakes
related talks
27mImplementing GitOps with ArgoCD: A Practical Guide
29mDemystifying Gitops with ArgoCD
46mProgressive Delivery in the Kubernetes era
30mOur Journey in Integrating GitOps Tools Into Delivery Pipelines
21mContinuous Delivery: The Missing Piece in the GitOps-OCI Security Puzzle
43mGitOps at Scale with and Argo CD on
20mMulti-Stage Deployment Pipelines the GitOps Way
16mUse Kubernetes-native Integrated GitOps CI/CD Workflows