| name | design-deployment-strategy |
| description | Use when selecting or designing a deployment strategy for releasing software to production safely |
| source | AWS Well-Architected Framework (Reliability Pillar); Netflix Tech Blog (canary deployments); Martin Fowler "BlueGreenDeployment" (martinfowler.com) |
| tags | ["deployment","blue-green","canary","rolling","devops","reliability","aws"] |
| verified | true |
Design Deployment Strategy
Select and implement the deployment strategy that best matches the system's risk tolerance, rollback requirements, and infrastructure capabilities.
Why This Is Best Practice
Adopted by: Netflix (canary), AWS (blue-green with CodeDeploy), Google (gradual rollouts in GKE), Facebook (progressive push)
Impact: Netflix's canary deployments catch ~95% of production issues before they affect all users; blue-green deployments reduce mean time to recover (MTTR) from hours to minutes via instant rollback.
Why best: The choice of deployment strategy directly determines blast radius when something goes wrong. A rolling deploy with no traffic control can expose 100% of users to a bad release in minutes; canary deploys can limit exposure to 1% while metrics are evaluated.
Steps
- — High-risk changes (DB migrations, major refactors): blue-green or canary. Low-risk patches: rolling. Zero-downtime requirement: any strategy except in-place restart.
