Comprendre les déploiements Kubernetes

Welcome to the world of Kubernetes Deployments! After learning about Pods, you might wonder how to manage many identical Pods reliably.

Deployments are a powerful Kubernetes resource that helps you declare and manage the desired state of your applications.

Imagine running a web server. You need multiple copies (replicas) for high availability and to handle traffic. If one crashes, you need another to take its place.

Deployments automate these tasks:

• Self-Healing: Replaces failed Pods automatically.
• Scaling: Easily adjust the number of Pod replicas.
• Updates: Perform rolling updates without downtime.
• Rollbacks: Revert to previous versions if something goes wrong.

The core concept behind a Deployment is its desired state. You tell Kubernetes what you want your application to look like, and it works to make it happen.

For example, if you say "I want 3 Nginx Pods running version 1.21," the Deployment will ensure exactly 3 Pods with Nginx 1.21 are running, even if some fail.

Under the hood, Deployments use another Kubernetes resource called a ReplicaSet.

• A ReplicaSet ensures a specified number of Pod replicas are running at all times.
• When you create a Deployment, it automatically creates a ReplicaSet.
• When you update a Deployment, it creates a new ReplicaSet and gracefully scales down the old one while scaling up the new one.

Like most Kubernetes resources, Deployments are defined using YAML. Key fields include:

• apiVersion: apps/v1 (for Deployments)
• kind: Deployment
• metadata: Name, labels to identify the Deployment.
• spec:
• replicas: The desired number of Pods.
• selector: How the Deployment finds its Pods (using labels).
• template: The Pod definition (just like a standalone Pod YAML!).

Let's create a simple Deployment for Nginx. This YAML specifies 3 replicas of an Nginx web server.

You would typically save this as nginx-deployment.yaml and apply it using kubectl apply -f nginx-deployment.yaml.

Once you've applied your Deployment, you can check its status using kubectl commands:

• kubectl get deployments: Shows a summary of your deployments.
• kubectl get pods -l app=nginx: See the Pods managed by your Nginx deployment.
• kubectl describe deployment nginx-deployment: Get detailed information, including events and associated ReplicaSets.

One of the best features of Deployments is rolling updates. This allows you to update your application version without downtime.

When you change the image or configuration in your Deployment's YAML, Kubernetes:

1. Creates new Pods with the updated version.
2. Gradually terminates old Pods as new ones become ready.

This ensures your application remains available throughout the update process.

Let's update our Nginx Deployment to a newer version. You would edit your nginx-deployment.yaml file and change the image to nginx:1.22.1.

Then, apply the changes: kubectl apply -f nginx-deployment.yaml. Kubernetes will perform a rolling update!

Sometimes an update goes wrong. Deployments allow you to easily rollback to a previous stable version.

Kubernetes keeps a history of your Deployments' revisions. You can see this history and revert to an earlier state if needed, minimizing the impact of bad deployments.

• kubectl rollout history deployment nginx-deployment: View past revisions.
• kubectl rollout undo deployment nginx-deployment --to-revision=1: Revert to a specific revision.

Congratulations! You've successfully understood Kubernetes Deployments.

We covered:

• What Deployments are and why they are crucial for managing applications.
• The concept of desired state and how Deployments use ReplicaSets.
• How to create, inspect, update, and rollback Deployments.

Next, we'll explore how to make your applications accessible using Kubernetes Services!