Introcktion

This article is the second in a series demonstrating different deployment strategies for modern cloud‑native applications. Following our previous exploration, this time we will focus specifically on Blue‑Green Deployment, and show how to implement it effectively using Linkerd as the service mesh.

This is a direct continuation of the first article:

Linkerd + Deployment Strategy – PART 1 {Canary}

The whole series was originally inspired by a lively discussion among fellow DevOps engineers in a WhatsApp group, where we compared and debated the best approaches for safe, reliable software releases.

In Part 1, we successfully walked through setting up and running Canary Deployments using the Linkerd service mesh. You can refer back to that guide, as well as access all the configuration files used, via the links below:

🔗 First Article: Linkerd + Deployment Strategy – PART 1 {Canary}

📂 GitHub Repository: Linkerd+Canary

What is Blue‑Green Deployment?

Blue‑Green Deployment is a release strategy designed to eliminate downtime and reduce risk when rolling out new versions of an application. It works by maintaining two identical production environments — referred to as Blue (the current stable version) and Green (the new version being released). At any given time, only one environment serves live production traffic. Once the Green environment is fully tested and validated, traffic is switched over from Blue to Green in a single, controlled action. If any issues arise, switching back to the Blue environment is just as fast and simple.

Why use Linkerd for this?

Linkerd is a lightweight, open‑source service mesh that adds reliability, security, and observability to Kubernetes applications without requiring changes to your code. It provides built‑in traffic management capabilities that make Blue‑Green deployments straightforward:

• Precise control over how traffic is routed between versions
• Real‑time metrics and health checks to monitor performance
• Instant rollback if problems are detected
• No complex custom scripting required

In this guide, we will walk through the full setup: preparing both Blue and Green deployments, configuring Linkerd to manage traffic between them, performing the switchover, and verifying the results.

So lets get to it.

I will skip the installation steps for Linkerd, as we are continuing directly from where we left off in PART 1. In the first part, we explicitly stated that we would use Minikube as our base environment and architecture. For this guide, it is assumed that:

• Minikube is up and running
• Linkerd has already been successfully installed and integrated into your cluster

List Working Files

Once again we are going to be listing the YAML files we are going to be working with

• namespace.yaml
• blue-deployment.yaml
• green-deployment.yaml
• service.yaml
• httproute.yaml
• deploy.sh