What is Blue-Green Deployment?

What is Blue-Green Deployment?

Blue-green deployment runs a new release beside the current production environment and then switches traffic when the new environment is ready. Blue usually means the live version and green means the candidate version, but the names are just labels; the key point is having two production-capable environments.

The new version is deployed with real configuration but without user traffic. Health checks, smoke tests, and critical user journeys run first. If they pass, a load balancer, reverse proxy, DNS change, or Kubernetes ingress rule sends traffic to the new environment.

When Is It Useful?

Blue-green deployment fits web applications, APIs, and internal portals where downtime is costly. If the release misbehaves, traffic can be pointed back to the previous environment quickly, making rollback faster than deploying a fresh package.

The tradeoff is cost and operational discipline. Two environments may run at the same time, and database schema changes must remain backward compatible if rollback is expected to work. Monitoring, migration planning, and feature flags are usually part of the release design.

CI/CD pipelines can automate the checks and switch. In Kubernetes, the same idea often appears as a service selector or ingress change.