Decoupling Canary Deployments From DBs With Argo Rollouts
summary
Shows how to decouple schema changes from deployments using Argo Rollouts by making each migration backwards and forwards compatible so old and new versions coexist. Uses Argo Rollouts CLI and a three‑phase pattern: phase 1 add columns, then deploy a canary that dual‑writes while reading new columns, then deploy the final app and phase 3 drop old column. Live demo renaming full_name → first/last succeeds with safe rollbacks until the final drop, which is irreversible.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:10Progressive delivery explained
- 2:54Challenges with stateful applications and databases
- 6:03DB refactoring patterns
- 7:51Multi-stage rollouts
- 12:06Argo Rollouts demo
- 17:41Key takeaways
- 19:45Supporting the deaf and hard of hearing community
- 21:50Q&A
related talks
25mJuggling Argo Rollouts for Progressive Delivery Across Multiple-Services
24mArgo Rollouts from the Trenches – Next Level Progressive Delivery Best Practices
30mArgo Rollouts at Scale: How We Brought Automated Rollback to 2100+ Micro-serv...
18mTailored Deployment Strategies: Argo Rollouts for Diverse Applications
29mTuning Argo Rollouts for Thousands of Workloads
01h 27mProgressive Delivery with Argo Rollouts
06m⚡ Lightning Talk: Improving Release Reliability with Argo Rollouts Across 10k Customer Services
40mProgressive Delivery with Kubernetes and Argo Rollouts