AWS vs GCP vs Azure: Cloud Platform Comparison

AWS vs GCP vs Azure

Choosing the right cloud platform for your project.

AWS

Largest ecosystem, most services, steepest learning curve.

#### EC2 Instance Lifecycle

EC2 instances follow a defined lifecycle: pending → running → rebooting → stopping → stopped → shutting-down → terminated. Understanding these states is key to managing costs and availability.

#### AMI Options

When launching an EC2 instance you choose an Amazon Machine Image (AMI):

• **Amazon Linux 2** — recommended default, comes with pre-installed AWS tools and tight integration with AWS services.
• **Ubuntu** — popular for multi-cloud workloads and memory-heavy applications; large community support.
• **SUSE Linux Enterprise** — aimed at enterprise environments, ships with a pre-installed LAMP stack.

#### EC2 Setup Checklist

1. Select the appropriate AMI for your workload. 2. Configure a VPC with public and private subnets. 3. Enable auto-assign public IP on the public subnet (or attach an Elastic IP for a static address). 4. Create security groups allowing SSH (port 22) and HTTP (port 80) inbound traffic. 5. Add user data scripts for automated bootstrapping on first launch. 6. Generate and securely store a .pem key pair for SSH access.

#### VPC Architecture

A typical production VPC flow looks like: Client → Internet Gateway → Router → Network ACL → NAT Gateway → Security Groups → Public Subnet (EC2) → Private Subnet (RDS). This layered approach ensures defense-in-depth while keeping databases off the public internet.

#### NAT Gateway

The NAT Gateway must live in a PUBLIC subnet and requires an Elastic IP. It translates private IP addresses so that instances in private subnets can reach the internet for updates and API calls, but external traffic cannot initiate connections inward — it is strictly one-directional outbound.

#### EC2 Billing

Linux and Windows instances are billed per second with a 60-second minimum. All other operating systems are billed per hour with a 1-hour minimum.

#### AWS Free Tier Highlights

• **S3**: 5 GB storage, 20,000 GET requests, 2,000 PUT requests per month.
• **CloudFront**: 50 GB data transfer out and 2 million HTTP/HTTPS requests per month.

GCP

Best for AI/ML, Kubernetes, and BigQuery analytics.

Azure

Best for enterprises, .NET, and Microsoft integration.

Recommendations

Startups: AWS or GCP. Enterprise: Azure or AWS. AI/ML: GCP.