My Children Will Never Deploy Active-Passive
summary
Argues that active‑active regional deployments beat active‑passive for resilience and cost, and shows a concrete stack: Terraform provisioning, Azure Spring Apps with Cloud Native Buildpacks and Spring Boot actuators for health/metrics, Azure Front Door global routing, and Redis Enterprise configured active‑active multi‑region. Demo deploys identical apps in Central US and South Central, writes replicate across regions (typical cross‑region latency ~15 ms, multi‑region up to five), and deleting a region leaves the app serving traffic.
AI-written summary · about
chapters (13)
- 0:00Introduction and background
- 3:37Understanding zones and regions
- 5:18Problems with active-passive setups
- 8:08Infrastructure as code and GitOps
- 12:26Moving from active-passive to active-active
- 18:21Scaling data and connection pooling
- 21:53Data replication strategies
- 26:07Demo setup with Terraform
- 32:51Global load balancing with Azure Front Door
- 36:44Scaling Kubernetes clusters across regions
- 41:08Deploying Redis Enterprise for resilience
- 47:03Deleting a region and testing resilience
- 48:22Q&A on data store replication
related talks
51mAchieving Resiliency With A Multi-Cloud Or Multi-Region Approach
18mLeveraging Multiple Regions to Improve Site Reliability:
45mMulti-Kubernetes, Multi-Region — Help!
22mCloud-Native Apps, recovery ready!
33mAvailability, Latency and Cost: Withstanding Regional Outages
25mChaos Engineering for Cloud native Apps
29mMulti Region Cloud Architecture for Data Residency Compliance
27mFailover Without Fear: A Practical Guide to Active-Active...