Portfolio

← Back to Blog

Deploy an Application to Google Kubernetes Engine (GKE)

Official path to deploy a containerized app on GKE: create cluster, build image, deploy with kubectl or Cloud Deploy. Prerequisites and links to Google Cloud docs only.

• 4 min read • By Jatinder (Jay) Bhola
gcpgoogle-cloudgkekubernetescontainerstutorialcertification

Why GKE for the Professional Cloud Developer

Google Kubernetes Engine (GKE) is Google Cloud’s managed Kubernetes service. The Professional Cloud Developer certification expects you to understand how to deploy and run applications on GCP, including containerized workloads. GKE is the primary option when you need orchestration, multi-service architecture, or fine-grained control over scaling and networking.

This post summarizes the official Google Cloud deployment paths for GKE. All links are to Google sources only: Deploy an app to a GKE cluster, Cloud Deploy for GKE, and related docs.

Deployment options: direct GKE vs Cloud Deploy pipeline

Two Main Deployment Approaches

Approach Best for Tools
Direct GKE deployment Learning, single cluster, simple pipelines gcloud, kubectl, Cloud Build (optional)
Cloud Deploy Multi-environment, promotion, delivery pipelines Cloud Deploy, Skaffold, Kubernetes manifests

Prerequisites (Official Checklist)

Before deploying to GKE you need:

  1. Google Cloud project with billing enabled.
  2. APIs enabled:
    • Kubernetes Engine
    • Artifact Registry (for container images)
    • Cloud Build (if you build images in GCP)
      For Cloud Deploy: Cloud Deploy quickstart lists the exact APIs.
  3. IAM: At least sufficient permissions to create clusters and deploy (e.g. Kubernetes Engine Developer or similar; see GKE access control).
  4. A container image — In Artifact Registry or another registry GKE can pull from.
  5. Kubernetes manifests — Deployment, Service, and optionally Ingress or ConfigMaps/Secrets.

Path 1: Deploy an App to a GKE Cluster (Direct)

Official doc: Deploy an app to a GKE cluster.

High-level steps:

  1. Create a GKE cluster (e.g. with gcloud container clusters create or the console).
  2. Get credentials for kubectl:
    gcloud container clusters get-credentials CLUSTER_NAME --region REGION
  3. Build and push your container image to Artifact Registry (or use a public image).
  4. Deploy using kubectl apply with your Deployment and Service YAML (or use a Helm chart if your org uses Helm).
  5. Expose the app (e.g. LoadBalancer Service or Ingress) and verify.

The same doc covers deploying a sample app and rolling updates. For “create and deploy a containerized web app” in one flow, see also Cloud Shell: Deploy GKE app (Google Cloud).

Path 2: Deploy with Cloud Deploy (Pipeline)

When you want repeatable pipelines and promotion (e.g. dev → staging → prod):

  • Quickstart: Deploy an app to GKE using Cloud Deploy.
  • You define delivery pipelines, targets (GKE clusters), and Skaffold configuration so Cloud Deploy can render manifests and deploy.
  • Skaffold is used to build images and apply Kubernetes manifests; the quickstart walks you through the required config.

Key Concepts for the Exam

  • Cluster — Control plane + node pool(s); you manage workloads, Google manages the control plane.
  • Deployment — Declarative way to run a set of Pods (replicas, image, env, resources).
  • Service — Stable network endpoint (ClusterIP, LoadBalancer, NodePort) for Pods.
  • Image registry — Artifact Registry is the recommended place to store container images used by GKE.
  • Rollouts — Rolling updates, rollback; both kubectl and Cloud Deploy support them.

Where to Practice

References (Google sources only)