Guided Lab: Using Startup Scripts in GCP VM Instances

Jon-Bonso

Description

A startup script is a file that executes tasks when a virtual machine (VM) instance boots. It can be applied either to all VMs within a project or to a specific VM. If both project‑level and VM‑level scripts are defined, the VM‑level script takes precedence. Startup scripts run only when network connectivity is available.

In this lab, you will learn how to configure a Google Cloud Platform (GCP) VM instance to automatically run a startup script upon boot. The script will install and configure Nginx, serve a custom HTML page, and apply a basic server configuration.

Prerequisites

To ensure the successful completion of this lab, you must have prior experience in creating VM instances and be familiar with their essential components. If you feel that your knowledge in this area is insufficient, we highly recommend taking this lab to gain the necessary understanding:

Guided Lab: How to Launch a GCP Compute Engine Linux Instance

Objectives

In this lab, you will:

Launch a Compute Engine VM instance with a custom startup script.
Automatically install and configure Nginx.
Serve a custom HTML page via HTTP.
Verify the deployment using the VM’s external IP address.
Inspect the VM’s metadata to confirm that the startup script is correctly applied.

Subscribe to access GCP PlayCloud Labs

Lab Steps

Create a Compute Engine VM Instance

1. Create a VM instance using the following machine configurations:

Name: Enter your desired instance name.
Machine type: Choose e2-micro (Preset)

2. Under OS and storage, click the Change button to modify the Boot disk:

Operating System: Ubuntu
Version: 24.04 LTS Minimal (x86/64, amd64 noble minimal image built on 2025‑12‑17)
Click Select once done.

3. Under Network → Firewall, check:

Allow HTTP traffic

4. Navigate to the Advanced section.

5. In the Startup script field, paste the following script:

# Update system
sudo apt update -y

# Install Nginx
sudo apt install nginx -y

# Start Nginx
sudo systemctl start nginx

# Create a custom HTML page in the HTML directory
echo '<h1>Welcome to my web page!</h1>' | sudo tee /usr/share/nginx/html/mypage.html > /dev/null

# Add the configuration directly
echo 'server {
    listen 80 default_server;
    server_name _;
    root /usr/share/nginx/html;
    
    location / {
        index mypage.html;
    }
}' | sudo tee /etc/nginx/conf.d/server.conf > /dev/null

# Remove the default site to avoid duplicate default_server errors:
sudo unlink /etc/nginx/sites-enabled/default

# Test and Reload Nginx for the changes to take effect
sudo nginx -t && sudo service nginx reload

# Enable Nginx at boot
sudo systemctl enable nginx

6. Leave other configurations at their default values.

7. Once done, click the Create button.

Wait for the VM to finish provisioning.

Verify the Setup

1. Once the VM is running, copy its External IP address.

2. Open a browser and navigate to http://<EXTERNAL_IP>.

3. You should see the custom page: “Welcome to my web page!”

If the page does not load:

Refresh the browser.
Ensure the firewall rule Allow HTTP traffic is enabled.
SSH into the VM and run: systemctl status nginx to confirm Nginx is running.

Verify the Startup Script via Metadata

1. Connect to your VM using SSH from the Cloud Console.

2. Run the following command to retrieve the startup script stored in the VM’s metadata:

curl -H "Metadata-Flavor: Google" \
  http://metadata.google.internal/computeMetadata/v1/instance/attributes/startup-script

Congratulations! You have automated the setup of a VM with a startup script, seamlessly installed and configured Nginx, and deployed a custom HTML page over HTTP. This exercise highlights how startup scripts can simplify server provisioning in GCP by ensuring services are installed, configured, and ready to run the moment your instance boots.