On this page

EC2 Compute

Checking access...

Amazon Elastic Compute Cloud (EC2) is the core compute service on AWS. It provides resizable virtual machines (instances) that you can launch within minutes.

Instance Types

EC2 instances are categorized into families optimized for different workloads:

Family	Use Case	Example Types
General purpose	Web servers, dev environments	t3, t4g, m7i
Compute optimized	Batch processing, HPC	c7i, c6g
Memory optimized	Databases, in-memory caches	r7i, x2iedn
Storage optimized	Big data, data warehousing	i4i, d3
Accelerated computing	ML, GPU rendering	p5, g5, trn1

Instance types follow a naming convention: {family}{generation}.{size}, for example t3.large or m7i.xlarge.

Tip

Use t4g instances (Graviton ARM processors) for cost savings — they offer up to 20% better price-performance compared to x86 equivalents for many workloads.

Amazon Machine Images (AMIs)

An AMI is a template that contains the OS and software configuration for an instance. AMIs are regional and identified by an ID (ami-xxxxxxxx).

Common AMI sources:

AWS-provided — Amazon Linux 2, Ubuntu, Windows Server, RHEL, SUSE
AWS Marketplace — Third-party AMIs with pre-installed software
Custom AMIs — Created from running instances using aws ec2 create-image

# Search for available AMIs
aws ec2 describe-images --owners amazon --filters "Name=name,Values=amzn2-ami-hvm-*-x86_64-gp2" --query 'Images[*].[ImageId,Name,CreationDate]' --output table

Security Groups

Security groups act as virtual firewalls for EC2 instances. They are stateful — if you allow inbound traffic, the outbound response is automatically allowed.

resource "aws_security_group" "web_sg" {
  name        = "web-server-sg"
  description = "Allow HTTP and SSH"

  ingress {
    from_port   = 80
    to_port     = 80
    protocol    = "tcp"
    cidr_blocks = ["0.0.0.0/0"]
  }

  ingress {
    from_port   = 22
    to_port     = 22
    protocol    = "tcp"
    cidr_blocks = ["10.0.0.0/8"]
  }

  egress {
    from_port   = 0
    to_port     = 0
    protocol    = "-1"
    cidr_blocks = ["0.0.0.0/0"]
  }
}

Caution

Security group rules are evaluated as a whole — there is no DENY rule within a security group. To block specific traffic, use Network ACLs (stateless) at the subnet level.

Pricing Models

AWS offers four purchasing options for EC2:

Model	Description	Savings vs On-Demand	Best For
On-Demand	Pay per hour/second, no commitment	None	Short-term, unpredictable workloads
Reserved	1 or 3 year commitment, partial/upfront	Up to 72%	Steady-state production workloads
Spot	Bid on spare capacity, can be interrupted	Up to 90%	Fault-tolerant, stateless workloads
Savings Plans	Flexible compute commitment ($/hour)	Up to 72%	Consistent compute across EC2, Fargate, Lambda

# Request a Spot instance
aws ec2 request-spot-instances \
  --spot-price "0.05" \
  --instance-count 1 \
  --type "one-time" \
  --launch-specification "{\"ImageId\":\"ami-0c55b159cbfafe1f0\",\"InstanceType\":\"t3.micro\"}"

Placement Groups

Placement groups control how instances are placed on underlying hardware:

Type	Strategy	Use Case
Cluster	Low-latency, same rack	HPC, tightly coupled workloads
Spread	Distinct hardware	Critical instances, fault isolation
Partition	Groups across partitions	Distributed systems (Hadoop, Cassandra)

resource "aws_placement_group" "cluster" {
  name     = "hpc-cluster"
  strategy = "cluster"
}

User Data and Bootstrapping

User data scripts run at instance launch to automate configuration:

#!/bin/bash
yum update -y
yum install -y httpd
systemctl start httpd
systemctl enable httpd
echo "<h1>Hello from $(hostname -f)</h1>" > /var/www/html/index.html

Key Takeaways

EC2 instance types follow {family}{gen}.{size} — choose based on workload: general purpose (t/m), compute (c), memory (r/x), storage (i/d), accelerated (p/g/trn)
Security groups are stateful firewalls at the instance level — inbound rules control traffic, outbound responses are auto-allowed; no explicit DENY rules
Four pricing models: On-Demand (flexible), Reserved (steady-state, up to 72% off), Spot (fault-tolerant, up to 90% off), Savings Plans (compute-agnostic commitment)
AMIs are regional templates — source from AWS, Marketplace, or create custom images with aws ec2 create-image
Placement groups control hardware placement: Cluster (low-latency same rack), Spread (distinct hardware), Partition (groups across racks)
User data scripts bootstrap instances at launch using #!/bin/bash — ideal for installing packages and configuring services