deployment-patterns

Rolling deployment updates instances gradually, blue-green switches traffic between two identical environments, and canary releases to a subset of users first. Each strategy offers distinct trade-offs in speed, risk, and rollback complexity for production releases.

Multi-stage Dockerfiles use separate builder and runtime stages to compile your application and copy only the final artifacts into a minimal base image. This pattern reduces final image size and improves security for production deployments.

Kubernetes probes check container health using liveness and readiness endpoints. Liveness probes restart failing containers, while readiness probes control traffic routing to ensure only healthy pods receive requests during deployments.

A production readiness checklist verifies health checks, rollback strategies, environment configuration, and CI/CD pipeline stages. It ensures applications follow the Twelve-Factor App pattern for configuration and have proper Docker containerization before release.

GitHub Actions CI/CD pipelines automate deployment through defined workflow stages that build, test, and push Docker images. Rollback strategies execute by re-deploying previous stable image tags when health checks fail in production.

Use canary deployment when you need to validate changes with a small subset of traffic before full release, reducing risk. Choose rolling updates for faster, simpler deployments when you have high confidence in your application's backward compatibility.