Cloud Shell (AWS & GCP)

5 minute read

Cloud shells give you a full Linux terminal in your web browser — no installation, no configuration. They are perfect for:

  • Learning Linux when you cannot install software locally
  • Running cloud CLI commands (aws, gcloud, az) without configuring your machine
  • Quick experiments and one-off tasks
  • Using Linux from any device (tablet, library computer, etc.)

AWS CloudShell

AWS CloudShell is a free browser-based terminal pre-configured with the AWS CLI.

Getting Started

  1. Sign in to the AWS Console
  2. Click the CloudShell icon in the top navigation bar (terminal icon)
  3. Or go to: https://console.aws.amazon.com/cloudshell/

AWS CloudShell opens an Amazon Linux 2023 terminal with your AWS credentials automatically loaded.

What’s Pre-installed

# AWS CLI (with your credentials automatically available)
aws --version
# aws-cli/2.x.x Python/3.x Amazon Linux

# Your identity is automatically configured
aws sts get-caller-identity
# {
#     "UserId": "AIDA...",
#     "Account": "123456789012",
#     "Arn": "arn:aws:iam::123456789012:user/your-username"
# }

# Common tools
python3 --version   # Python 3.9+
node --version      # Node.js 18+
git --version       # git 2.x
jq --version        # jq 1.6

Installing Additional Tools

# Install packages with dnf (Amazon Linux 2023 uses dnf)
sudo dnf install -y htop tree nmap

# Install Python packages
pip3 install boto3 requests

# Install Node packages globally
npm install -g serve

# Download binaries directly
curl -LO https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/basics/amd64/kubectl
chmod +x kubectl
sudo mv kubectl /usr/local/bin/
kubectl version --client

Storage and Persistence

# CloudShell gives you 1 GB of persistent storage in your home directory
df -h ~
# Filesystem  Size  Used  Avail  Use%  Mounted on
# overlay     1.0G  312M  712M   31%   /home/cloudshell-user

# Files in ~ persist between sessions
# Files outside ~ are lost when the session ends

# Practical: store scripts and configs in ~/
mkdir -p ~/scripts ~/configs
echo '#!/usr/bin/env bash
echo "Hello from CloudShell"' > ~/scripts/hello.sh
chmod +x ~/scripts/hello.sh

Limitations

Limitation Detail
Idle timeout Session disconnects after ~20 minutes idle
Storage 1 GB persistent home directory
Network Outbound only; cannot receive incoming connections
Region CloudShell session is region-specific
No root sudo works but some system changes don’t persist
Compute Limited CPU/RAM (shared environment)

Google Cloud Shell

Google Cloud Shell is the most generous free cloud terminal available — 5 GB persistent home directory and a real VM.

Getting Started

  1. Sign in to https://console.cloud.google.com/
  2. Click the Activate Cloud Shell icon (terminal icon in the top-right)
  3. Or go directly to: https://shell.cloud.google.com/

Google Cloud Shell runs on a Debian Linux VM with your Google Cloud credentials pre-configured.

What’s Pre-installed

# Check the environment
uname -a
# Linux cs-... Debian GNU/Linux

# Google Cloud SDK (gcloud)
gcloud --version
# Google Cloud SDK 465.0.0

# Your identity is automatically configured
gcloud auth list
# ACTIVE  ACCOUNT
# *       your-email@gmail.com

# Development tools
python3 --version   # Python 3.11+
node --version      # Node.js 20+
java --version      # OpenJDK 11
go version          # Go 1.21+
docker --version    # Docker (Cloud Shell only)
kubectl version     # kubectl
terraform --version # Terraform
git --version       # git

Boost Mode

For compute-intensive tasks, enable Boost Mode:

  • Click the three-dot menu in Cloud Shell → Enable Boost Mode
  • Provides a higher-spec VM (more CPU/RAM)
  • Standard: 1 vCPU, 1.7 GB RAM
  • Boost: 6 vCPU, 45 GB RAM (limited time)

5 GB Persistent Home Directory

# Your home directory persists indefinitely
df -h ~
# Filesystem  Size  Used  Avail  Use%  Mounted on
# /dev/sda1   5.0G  1.2G  3.8G   24%  /home/your-username

# Clone a GitHub repo — it persists between sessions
git clone https://github.com/your-org/your-project.git ~/projects/myproject

# Install tools in ~ — they persist
mkdir -p ~/bin
curl -LO https://github.com/mikefarah/yq/releases/latest/download/yq_linux_amd64
mv yq_linux_amd64 ~/bin/yq
chmod +x ~/bin/yq
echo 'export PATH="$HOME/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc

Web Preview

Cloud Shell lets you preview web applications:

# Start a web server on port 8080
python3 -m http.server 8080

# In Cloud Shell, click "Web Preview" → "Preview on port 8080"
# A browser window opens showing your server's response
# Accessible from anywhere — great for demos

Limitations

Limitation Detail
Idle timeout VM spins down after 1 hour idle
Free hours 50 hours/week of Boost Mode; Standard is always free
Storage 5 GB persistent (data deleted if inactive 120 days)
No persistent VMs VM is ephemeral (home dir persists, VM does not)

Azure Cloud Shell

Azure Cloud Shell provides both Bash and PowerShell in your browser.

Getting Started

  1. Sign in to https://portal.azure.com/
  2. Click the Cloud Shell icon in the top navigation bar
  3. Or go to: https://shell.azure.com/

The first time, you’ll be prompted to create a storage account (required for file persistence).

# Choose Bash when prompted (not PowerShell)

# Azure CLI is pre-configured with your identity
az account show
# {
#   "name": "My Subscription",
#   "user": { "name": "you@example.com" }
# }

# Tools available
az --version
kubectl version --client
terraform --version
git --version
python3 --version

Storage

Azure Cloud Shell uses a mounted Azure File Share for persistence:

# Your persistent storage
df -h ~
# 5 GB Azure Files storage mounted at ~/
# Billed as standard Azure storage (very cheap — ~$0.10/GB/month)

# Files persist indefinitely between sessions

Comparing the Three Cloud Shells

Feature AWS CloudShell Google Cloud Shell Azure Cloud Shell
Cost Free Free Free (storage ~$0.05/mo)
Persistent storage 1 GB 5 GB 5 GB (Azure Files)
OS Amazon Linux 2023 Debian Linux Ubuntu/Debian
Idle timeout 20 min 60 min 20 min
Pre-installed CLIs aws gcloud, kubectl, terraform az, kubectl, terraform
Docker No Yes (limited) No
Web preview No Yes Yes
Boost/Ephemeral VM No Yes (6 vCPU) No
Best for AWS practice Learning Linux + GCP Azure practice

Practical Workflow: Learning Linux in the Cloud

# Recommended: Use Google Cloud Shell for learning Linux

# 1. Open https://shell.cloud.google.com
# 2. Create a dedicated learning folder
mkdir -p ~/linux-learning/scripts
cd ~/linux-learning

# 3. Create and run your first shell script
cat > ~/linux-learning/hello.sh << 'EOF'
#!/usr/bin/env bash
echo "Welcome to Linux!"
echo "Today is: $(date)"
echo "You are: $(whoami)"
echo "Your home is: $HOME"
echo "Your shell is: $SHELL"
EOF

chmod +x ~/linux-learning/hello.sh
./hello.sh
# Welcome to Linux!
# Today is: Tue Apr  1 10:30:00 UTC 2025
# You are: your-username
# Your home is: /home/your-username
# Your shell is: /bin/bash

# 4. Your script is saved — close the browser, reopen, it's still there

Key Takeaways

  • AWS CloudShell: free, 1 GB storage, pre-configured with aws CLI, 20-minute idle timeout
  • Google Cloud Shell: free, 5 GB storage, best for learning Linux, 1-hour idle timeout, web preview
  • Azure Cloud Shell: free (tiny storage cost), 5 GB storage, pre-configured with az CLI
  • All three run genuine Linux — the commands you learn work on any Linux server
  • Google Cloud Shell is the best choice for pure Linux learning (most storage, longest timeout, most pre-installed tools)
← Back to macOS Terminal Setup Next → First Login and Prompt

Updated:

You may also enjoy