Preparing for disaster: going Multi-Region
summary
Claims that migrating to multi-region hits 99.99% uptime by combining specific AWS features rather than vague HA advice: implement cross-account and cross-region replication for S3 and secrets, use multi-region KMS keys and secret manager replication, wire regions with Transit Gateway/VPC peering plus Route53 or Global Accelerator ~6s failover, and replicate data with DynamoDB Global Tables or Aurora read‑replicas / RDS logical replication. CI/CD with Terraform deploys both regions. Result: higher cost and complexity but observable cross‑region failover and greatly reduced outage surface.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:28High availability vs fault tolerance
- 4:20Multi-region disaster recovery importance
- 6:28Security and data replication
- 8:13Networking for global services
- 10:47Compute and container strategies
- 12:14Data consistency challenges
- 15:09CI/CD across regions
- 17:20Observability in multi-region
- 18:24Conclusion and key takeaways
related talks
29mMulti-Regional High Availability in AWS Cloud
14mExpect the unexpected: Building resilience with AWS
41mResiliency and data residency
51mBuilding a multi-account, multi-runtime service-oriented architecture
47mBuilding Resilience
18mLeveraging Multiple Regions to Improve Site Reliability:
54mFive hybrid cloud backup and disaster recovery mistakes to avoid
45mMulti-Kubernetes, Multi-Region — Help!