Data Residency: the fun of storing data worldwide
summary
Concrete proposal: route by country at the gateway, validate and correct at the backend, accepting double computation to fail fast and prevent cross‑jurisdictional reads. Demo uses APISIX and ShardingSphere with two apps, two databases (Europe and USA schemas) deployed via two-region Docker Compose demo and OpenTelemetry tracing. Default gateway routing causes one mismatched read, backend recomputation detects owner country and redirects or corrects the query, so the pattern works with occasional slower redirect paths and added implementation complexity.
AI-written summary · about
chapters (14)
- 0:00Introduction and speaker intro
- 1:57What is data residency
- 7:31Data partitioning strategies
- 16:37Client vs server routing decisions
- 17:00Demo setup and tools
- 21:43Live architecture demonstration
- 30:08Q&A on complexity and validation
- 31:52API Gateway vs reverse proxy
- 37:14Q&A on API 6 and caching
- 40:29Organizational inertia and advice
- 41:11Q&A on data separation
- 43:25Data cleanup and documentation
- 45:38European clouds and regulations
- 50:03Final Q&A and closing
related talks
31mToo Far Too Furious: Multi-Region Cloud Applications
30mMulti-Region, Multi-Primary, and Eventually Consistent? You're Crazy
30mSo You Want to Geo-Partition Your Data... - Deepthi Sigireddi & Jiten Vaidya, Planetscale, Inc.
17mEnabling Compute and Data Sovereignty with Infrastructure-Level Data Spaces
52mTaking Storage for a Ride
27mThe journey of Data Mesh and Compliance in a Multi-Regional Data Lake
19mDataEngBytes 2023 - SYD-T1-04 - Yash Sharma
19mWhat Breaks When You Build AI Under Sovereignty Constraints