What Is a Blue-Green Deployment?

Two Servers, One Switch

Imagine you have two identical cars. You paint one blue and one green. The blue car is the one currently driving people around — it's doing the job. The green car is parked nearby, clean, and ready. Whenever you want to upgrade the car, you work on the green one first. When it's ready and tested, you swap drivers over to green. Blue sits and waits.

That's blue-green deployment. You keep two identical server environments running at the same time. The blue server handles live traffic with the current version of your app. The green server gets the new version loaded and tested. When green is ready, a router instantly switches all users over to it. If anything breaks, you switch back to blue in seconds.

This approach means users never see a "site is down for maintenance" page. The switch happens so fast — often in under a second — that no one notices.

Zero Downtime Changes Everything

For businesses that sell online, every minute of downtime costs money. A store that goes down for 5 minutes during a sale loses real revenue. Blue-green deployment means you can push updates anytime — even on Black Friday — without users seeing a broken page or losing items in their cart.

It also means developers can move faster. Without this safety net, teams have to schedule maintenance windows, warn users in advance, and rush the deployment under pressure. With blue-green, you just deploy, test on the live environment, and flip the switch.

The Four-Step Swap

Here's how blue-green deployment works in practice, step by step:

The key to making this work is a piece of software called a load balancer or router. It's the traffic cop that decides: should users go to the blue server or the green server? You configure it to point to one or the other. Flipping the switch is just changing that configuration.

The whole process can be automated. A CI/CD pipeline can run tests automatically, deploy to green, run more tests, then flip the switch — all without a human touching anything.

A Simple Router Switch

Here's a very simple example of how a load balancer configuration decides which server gets traffic. Think of it like a toggle switch:

# Blue-Green Router Config — swap "active" between blue and green
environments:
  blue:
    url: https://blue.myapp.com
    status: standby
    version: v1.2.0

  green:
    url: https://green.myapp.com
    status: active
    version: v1.3.0

# The router sends all traffic to whichever has status: active
router:
  mode: blue_green
  active: green
  health_check_path: /health

# After deploying v1.3.0 to green and testing, flip active to green
# Green now serves all traffic. Blue sits as instant rollback option.

In a real deployment pipeline (like using GitHub Actions + AWS or Kubernetes), this switch happens automatically once tests pass. The old server (blue) is left running in standby until you're 100% sure the new version is solid — then it gets updated to the next release's new code and becomes the new standby.

Knowledge Check

Test what you learned with this quick quiz.