Lab 01: Step-by-Step Guide

Prerequisites Check

Before starting, ensure you have:

# Check common tools (required for both providers)
terraform version    # Should be >= 1.5
kubectl version --client
helm version

GCP-Specific Prerequisites

# Check GCP tools
gcloud version

# Check GCP authentication
gcloud auth list
gcloud config get-value project

AWS-Specific Prerequisites

# Check AWS tools
aws --version

# Check AWS authentication
aws sts get-caller-identity
aws configure list

Step 1: Choose Your Provider

Edit terraform.tfvars and set:

cloud_provider = "gcp"  # or "aws"

Step 2: Configure Provider-Specific Settings

Option A: GCP Configuration

# Set your project
export PROJECT_ID="your-project-id"
gcloud config set project $PROJECT_ID

# Enable required APIs
gcloud services enable \
  container.googleapis.com \
  compute.googleapis.com \
  artifactregistry.googleapis.com \
  logging.googleapis.com \
  monitoring.googleapis.com

Edit terraform.tfvars:

cloud_provider = "gcp"
project_id     = "your-project-id"
cluster_name   = "implementation-studio"
region         = "us-central1"
machine_type   = "e2-medium"

Option B: AWS Configuration

# Verify AWS credentials
aws sts get-caller-identity

# Verify you have required permissions for EKS
aws iam get-role --role-name AWSServiceRoleForAmazonEKS || echo "Service-linked role will be created automatically"

Edit terraform.tfvars:

cloud_provider     = "aws"
cluster_name       = "implementation-studio"
region             = "us-west-2"
vpc_cidr          = "10.0.0.0/16"
availability_zones = ["us-west-2a", "us-west-2b"]
instance_type     = "t3.medium"

Step 3: Configure Terraform Variables

cd labs/01-standard-deployment
cp terraform.tfvars.example terraform.tfvars

Edit terraform.tfvars with your provider-specific values (see Step 2).

Step 4: Initialize Terraform

terraform init

This downloads the required providers (Google and/or AWS) and modules.

Step 5: Review Terraform Plan

terraform plan

Review the resources that will be created:

GCP:

• VPC network and subnets
• GKE cluster
• Artifact Registry

AWS:

• VPC network and subnets
• EKS cluster
• ECR repository

Step 6: Apply Infrastructure

terraform apply

This will take 10-15 minutes. The cluster creation is the longest step.

GCP: GKE cluster creation typically takes 5-10 minutes AWS: EKS cluster creation typically takes 10-15 minutes

Step 7: Get Cluster Credentials

After Terraform completes, get cluster credentials.

GCP

CLUSTER_NAME=$(terraform output -raw cluster_name)
REGION=$(terraform output -raw cluster_location)
PROJECT_ID=$(terraform output -raw project_id || grep project_id terraform.tfvars | cut -d'"' -f2)

gcloud container clusters get-credentials $CLUSTER_NAME \
  --region $REGION \
  --project $PROJECT_ID

AWS

CLUSTER_NAME=$(terraform output -raw cluster_name)
REGION=$(grep -E '^region\s*=' terraform.tfvars | sed 's/.*=\s*"\(.*\)".*/\1/' | tr -d ' ')

aws eks update-kubeconfig \
  --region $REGION \
  --name $CLUSTER_NAME

Or Use the Helper Script

# The deploy-argo.sh script handles credentials automatically
./scripts/deploy-argo.sh

Verify access:

kubectl get nodes

Step 8: Deploy Argo Workflows

./scripts/deploy-argo.sh

This script:

1. Detects your provider (GCP or AWS)
2. Gets cluster credentials automatically
3. Adds Helm repositories
4. Creates namespaces
5. Installs Ingress NGINX
6. Installs Argo Workflows

Step 9: Verify Deployment

./scripts/validate.sh

Check that all pods are running:

kubectl get pods -n argo
kubectl get pods -n ingress-nginx

Step 10: Get Ingress IP

kubectl get service ingress-nginx-controller -n ingress-nginx

Note the EXTERNAL-IP address (or HOSTNAME for AWS).

GCP: Shows an IP address AWS: Shows a hostname (e.g., a1b2c3d4e5f6g7h8.elb.us-west-2.amazonaws.com)

Step 11: Submit a Sample Workflow

kubectl apply -f manifests/sample-workflow.yaml

Check workflow status:

kubectl get workflows -n argo
kubectl describe workflow hello-world -n argo

View workflow logs:

kubectl logs workflow/hello-world -n argo

Step 12: Access Argo Workflows UI (Optional)

Option 1: Port Forward (Works for Both Providers)

kubectl port-forward -n argo svc/argo-workflows-server 2746:2746

Then access: https://localhost:2746

Option 2: Create Ingress (Provider-Specific)

GCP Example:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: argo-workflows-ui
  namespace: argo
spec:
  ingressClassName: nginx
  rules:
  - host: argo.example.com  # Replace with your domain or use IP
    http:
      paths:
      - path: /
        pathType: Prefix
        backend:
          service:
            name: argo-workflows-server
            port:
              number: 2746

AWS Example:

Same YAML works, but you'll get an AWS ELB hostname. You can also configure a custom domain.

Cleanup

When finished:

./scripts/cleanup.sh

Or manually:

terraform destroy

Important: Destroying may take 10-15 minutes. Be patient, especially when removing load balancers and clusters.

Troubleshooting

Provider-Specific Steps

See Troubleshooting Guide for provider-specific troubleshooting steps.

Quick Checks

# Check cluster status (GCP)
gcloud container clusters describe $CLUSTER_NAME --region $REGION --project $PROJECT_ID

# Check cluster status (AWS)
aws eks describe-cluster --name $CLUSTER_NAME --region $REGION

# Check nodes
kubectl get nodes -o wide

# Check all resources
kubectl get all -n argo
kubectl get all -n ingress-nginx