Implementaciones automatizadas con CD de Kubernetes

After building and testing your code, the next step is getting it to users. Continuous Deployment (CD) automates this entire process, from code commit to production deployment, without human intervention.

It's the natural evolution of Continuous Integration (CI), where every successful build that passes all tests is automatically released.

Manually deploying to Kubernetes can be complex and error-prone. Automating deployments brings significant advantages:

• Speed: New features reach users faster.
• Consistency: Deployments are identical every time.
• Reliability: Reduces human errors in complex K8s operations.
• Scalability: Manages deployments across many services and clusters.

A basic Continuous Deployment pipeline for Kubernetes usually follows these stages:

1. Code Commit: Developer pushes code to Git.
2. CI (Build & Test): Automated build, tests run, Docker image created.
3. Image Push: New Docker image pushed to a container registry (e.g., Docker Hub).
4. K8s Manifest Update: Deployment manifest updated with the new image tag.
5. Apply to Cluster: The updated manifest is applied to the Kubernetes cluster.

In Kubernetes, you update an application by changing its Deployment manifest. The most common change is updating the Docker image tag to a newer version.

Kubernetes then intelligently manages the transition from the old version to the new one, ensuring your application remains available.

Here's a simplified Kubernetes Deployment manifest. To update your application, you simply change the image tag (e.g., from v1.0.0 to v1.0.1) and apply the updated manifest.

Kubernetes uses a rolling update strategy by default. When you update a Deployment, it:

• Gradually replaces old Pods with new ones.
• Ensures a minimum number of Pods are always running.
• Avoids downtime by not taking all old Pods down simultaneously.

This ensures a smooth transition to your new application version.

For automated deployments, liveness and readiness probes are crucial:

• Liveness Probe: Tells Kubernetes if your application is still running. If it fails, Kubernetes restarts the Pod.
• Readiness Probe: Tells Kubernetes if your application is ready to serve traffic. New Pods won't receive traffic until this probe passes.

These prevent unhealthy versions from taking user requests.

Even with automation, issues can arise. Kubernetes tracks deployment history, allowing for easy rollbacks:

• If a new version is buggy, you can quickly revert to a previous, stable version.
• The command kubectl rollout undo deployment/my-app-deployment can revert to the last successful deployment.

This capability is a safety net for CD.

Many CI/CD tools can automate Kubernetes deployments. They often integrate by:

• Running kubectl commands in pipeline scripts.
• Using Kubernetes APIs directly.

Examples include Jenkins, GitLab CI/CD, GitHub Actions, and CircleCI. These tools orchestrate the steps from image build to K8s deployment.

In this lesson, we explored Continuous Deployment (CD) for Kubernetes. You learned:

• CD automates application releases.
• How K8s manages updates using rolling updates.
• The importance of probes and rollbacks.
• How CI/CD tools integrate with Kubernetes for automation.

Next, we'll dive deeper into GitOps, a powerful methodology for K8s CD.