Progressive Deployments with Consul Service Mesh
summary
Shows a concrete progressive delivery pipeline built with Consul service mesh, Nomad, Prometheus and Grafana that automatically canaries and promotes releases. Automation clones the known‑good as primary, launches a candidate, monitors Prometheus metrics every 30s and sends 10% initial traffic then ramps by 20% steps, promotes at success or rolls back after two consecutive failure checks. Demo runs real splits (90/10, 50/50) and succeeds and fails safely, with retries hiding user errors while automation reverts bad canaries.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 2:08Deployment history and problems
- 7:47What is progressive delivery
- 10:13Consul Service Mesh setup
- 15:31Traffic control with subsets
- 25:27Automated release process
- 39:13Handling failed deployments
- 42:43Retrying failed requests
- 45:09Conclusion and Q&A
related talks
27mLeveraging Service Meshes for A/B, Blue-Green, and Canary Deployments (aka "Progressi...
36mProgressive Delivery in the Kubernetes era
51mProgressive Delivery in the Kubernetes era
40mProgressive Delivery with Kubernetes and Argo Rollouts
33mProgressive Delivery Techniques with Flagger
20mProgressive Delivery with the Kubernetes Gateway API
21mMonitoring Nomad with Prometheus and Icinga
35mProgressive Delivery with Argo Rollouts