Continuous Deployments and Data Sovereignty: A Case Study
summary
Builds actual multi‑country deployments: GitHub + Semaphore run tests and auto‑push a single master to Heroku for US plus Elastic Beanstalk containers in Oregon, Sydney and Canada, which proves regional deployments work but introduce latency, replication costs and broken cross‑region social features. Setup requires nontrivial ops work and environment management; team later scopes a PII/PHI data service with AES256 encryption and feature flags (Flipper) to keep identifiers local while the app remains global.
AI-written summary · about
chapters (13)
- 0:00Start
- 1:30You may know me from...
- 2:41A Few Words On Data
- 4:29Health Insurance Portability and Accountability Act (HIPAA)
- 6:40Data Sovereignty
- 8:30Continuous Deployment
- 10:08The Occasional Myth of Convenience
- 10:43Back To Our Startup
- 12:46Option #1
- 15:31Implementation
- 24:26Findings
- 25:22Next steps
- 27:28Recommendations
related talks
36mContinuous Deployment with Rails
26mThe Many Ways to Deploy Continuously
49mApplications architecture spanning multiple clouds
47mDeploying Like a Boss: From 2 Servers to 20,000
01h 03mContinuous Security in the Cloud
01h 00mReliability and speed by 'push to main' mentality
24mContinuous Deployment at Scale
46mContinuous Deployment Workflows