Zero-Downtime Deployment on Kubernetes
summary
Shows a concrete zero‑downtime path: keep app pods on Kubernetes, add a green database, run a streaming job that uses change data capture to transform and replicate historical and live rows, then do a rolling upgrade to point pods at the green DB. Uses Hazelcast Jet for the stream job and Hazelcast IMDG / Spring Sessions for session replication. Demo runs on kind, migrates untouched data online, and succeeds end to end.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 0:45Hazelcast
- 1:15Why ZeroDowntime
- 3:18BlueGreen Deployment Model
- 4:47State and Set
- 5:49Rolling
- 6:25Session Replication
- 7:24Database Options
- 8:02Rolling Upgrade
- 9:53eCommerce Example
- 12:12ECommerce Use Case
- 15:27Decomposition
- 17:07Data Migration
- 18:35Data Capture
- 21:50Memory Data Grid
- 22:25Spring Sessions
- 22:37Hazel Coast Jets
- 23:55Demo
- 26:30Deployment
- 30:05Rest Call
- 32:00Forward Job
- 34:10Summary
- 34:50Outro
related talks
37mZero downtime Deployment on Kubernetes
35mZero-downtime deployment on Kubernetes
43mZero-downtime deployment with Kubernetes, Spring Boot and Flyway
40mStrategies for Zero Down Time, Frequent Deployments
40mZero-Downtime Java Enterprise Applications With Kubernetes
37mZero-downtime Deployment Strategies (...) [ENG]
31mZero downtime deployment with Kubernetes, Spring Boot and Flyway
49mZero-Downtime Deployment in Kubernetes & Istio