EC2(Elastic Compute Cloud)

Leave a Comment / By /

Back to: AWS-Basics-Advanced

0

What is Amazon EC2 (Elastic Compute Cloud)?

In cloud computing, Amazon EC2 (Elastic Compute Cloud) is one of the most important and widely used services on AWS.

Think of EC2 as your virtual server in the cloud. Instead of buying and managing physical computers, AWS allows you to rent virtual machines (instances) on-demand. You can choose the amount of CPU, memory, storage, and operating system based on your requirements.

EC2 makes it possible to run applications, host websites, process big data, or deploy AI models — all with full control, scalability, and pay-as-you-go pricing.

Why Use EC2?

  1. ⚡ Scalable – You can scale instances up or down anytime.

  2. 💸 Cost-Effective – Pay only for the compute you use.

  3. 🔒 Secure – Integrated with IAM roles, firewalls, and encryption.

  4. 🌍 Global – Launch instances in any AWS region or availability zone.

  5. 🧠 Customizable – Choose OS (Linux, Windows), instance size, and storage type.

Main Components of EC2

Component Description
Amazon Machine Image (AMI) A template that defines the OS, application server, and software used to launch an instance.
Instance Type Specifies the CPU, memory, and network capacity (e.g., t2.micro, m5.large).
EBS Volume (Elastic Block Store) Persistent storage attached to your instance (like a virtual hard drive).
Key Pair Used for secure login (SSH or RDP). Includes a public and private key.
Security Groups Acts as a virtual firewall — controls inbound and outbound traffic to your EC2 instance.
Elastic IP A static public IP address that can be attached to your instance.
VPC (Virtual Private Cloud) The private network in which your EC2 instance runs.
User Data Optional script that runs automatically when your instance launches (for auto-installing software, etc.).
Elastic Load Balancer (ELB) Distributes traffic evenly across multiple EC2 instances.
Auto Scaling Group (ASG) Automatically adjusts the number of running EC2 instances based on load or demand.

Steps to Create an EC2 Instance (Step-by-Step Guide)

Follow these simple steps to create your first EC2 instance in the AWS Management Console 👇

Step 1: Log in to AWS Console

Go to AWS Management Console and log in to your account.

Step 2: Open EC2 Dashboard

Search for “EC2” in the AWS console search bar and click on EC2 under “Compute Services”.

Step 3: Launch an Instance

Click on “Launch Instance” button.
You’ll now configure your virtual machine.

Step 4: Choose a Name and Tag

Give your instance a name — for example:
Name: MyWebServer

Tags help organize your resources in AWS.

Step 5: Choose an Amazon Machine Image (AMI)

Select the operating system you want:

  • Amazon Linux 2 (default and free tier eligible)

  • Ubuntu

  • Windows Server

Step 6: Choose Instance Type

Select your desired instance type.
For example:

  • t2.micro – Free tier eligible (1 vCPU, 1GB RAM)

  • t3.medium – For higher workloads

Step 7: Configure Key Pair (Login)

Choose Create New Key Pair → Download the .pem file (for Linux) or .ppk file (for Windows).
This key will be used to SSH or RDP into your instance securely.

Step 8: Configure Network Settings

Choose:

  • VPC: Default or custom

  • Subnet: Choose any availability zone

  • Auto-assign Public IP: Enable

  • Security Group: Create a new one and allow:

    • SSH (port 22) for Linux

    • RDP (port 3389) for Windows

    • HTTP (port 80) if you’re hosting a website

Step 9: Configure Storage

By default, you’ll get an 8 GB EBS volume. You can increase it based on your needs.

Step 10: Launch Instance

Click Launch Instance — AWS will start creating your EC2 instance.

You can check the status under EC2 → Instances → Instance State.

Step 11: Connect to Instance

Once your instance is running, connect to it:

  • For Linux:
    Use the following SSH command from your terminal:

    ssh -i my-key.pem ec2-user@<Public-IP>

EC2 Instance Pricing Models — Overview

AWS offers multiple pricing models for EC2 so you can choose the right balance between cost, flexibility, and commitment.
The main models are:

Pricing Model Description Best For Cost Efficiency
On-Demand Instances Pay per hour/second for compute with no long-term commitment. Short-term, unpredictable workloads. đź’° Moderate
Reserved Instances (RI) Commit to a specific instance type for 1 or 3 years to get discounts. Steady-state workloads. đź’°đź’° High
Savings Plans Flexible alternative to RIs; commit to a consistent spend ($/hour). Long-term, consistent usage. đź’°đź’° High
Spot Instances Bid for unused EC2 capacity at up to 90% discount. Fault-tolerant, flexible workloads (batch, ML, CI/CD). đź’°đź’°đź’° Very High
Dedicated Hosts/Instances Physically isolated servers for compliance or licensing needs. Regulated industries or BYOL (bring your own license). đź’° High
Capacity Reservations Reserve capacity in a specific AZ for immediate availability. Mission-critical workloads needing guaranteed capacity. đź’° Moderate

1. On-Demand Instances

Description

  • Pay only for compute time you use (per second or per hour).

  • No upfront cost, no long-term commitment.

 Example Use Case

  • Development, testing, or short-term projects.

  • Startups experimenting with workloads.

 Example

  • You run a t2.micro instance for 10 hours/day → pay only for those 10 hours.

2. Reserved Instances (RI)

 Description

  • Commit to a specific instance type, region, and OS for 1 or 3 years.

  • Get up to 75% discount compared to On-Demand.

 Types of RIs

Type Flexibility Description
Standard RI ❌ Fixed Best discounts, fixed instance type.
Convertible RI âś… Flexible Change instance type/family if needed.
Scheduled RI ⏰ Time-based Run only during specific time windows.

 Example Use Case

  • Long-term applications like databases, backend servers.

3. Savings Plans

 Description

  • Flexible alternative to RIs — commit to spend $/hour for 1 or 3 years.

  • Works across instance families, sizes, regions.

 Types of Savings Plans

Plan Type Flexibility Description
Compute Savings Plan âś… Most Flexible Apply to any instance, region, or even Lambda/Fargate.
EC2 Instance Savings Plan ⚙️ Moderate Locked to an instance family in a region.

 Example Use Case

  • Enterprises with predictable cloud spend but dynamic workloads.

4. Spot Instances

 Description

  • Purchase unused EC2 capacity at up to 90% discount.

  • AWS can terminate with a 2-minute notice when capacity is needed.

Example Use Case

  • Big data processing, CI/CD pipelines, machine learning training, simulations.

 Example

  • Launch 50 Spot instances for a batch job → if AWS needs capacity back, instances stop — you can resume later.

5. Dedicated Hosts / Instances

 Description

  • Dedicated Instances: Run on hardware dedicated to you.

  • Dedicated Hosts: Gives control of physical servers for license management.

 Example Use Case

  • Compliance (HIPAA, PCI), software licenses tied to physical hardware.

6. Capacity Reservations

 Description

  • Reserve capacity in a specific Availability Zone.

  • Combine with RIs or Savings Plans for discounts.

 Example Use Case

  • Critical applications that must launch anytime (e.g., payment systems).

Amazon EC2 Instance Types — Based on Use Case

Amazon EC2 offers a wide range of instance families, each optimized for a specific purpose like compute, memory, storage, or GPU-based workloads.

1. General Purpose Instances

Description

Balanced mix of compute, memory, and networking — ideal for most workloads.

 Best For

  • Web servers

  • Development and test environments

  • Small to medium databases

 Examples

Family Example Types Highlights
t-series t3, t4g, t2 Burst performance, cost-effective for low-traffic apps.
m-series m6i, m5, m7g Balanced for compute & memory. Great for app servers.

Example Use Case

A startup runs its Node.js web app on an m5.large instance for balanced cost and performance.

2. Compute Optimized Instances

Description

Provide high-performance processors for compute-intensive tasks.

 Best For

  • High-performance web servers

  • Batch processing

  • Scientific modeling

  • Gaming servers

 Examples

Family Example Types Highlights
c-series c6i, c7g, c5 High compute-to-memory ratio, ideal for CPU-heavy apps.

 Example Use Case

A company runs 3D rendering and high-performance APIs using c6i.xlarge instances.

3. Memory Optimized Instances

 Description

Designed for memory-intensive workloads needing fast performance for large datasets.

 Best For

  • In-memory databases (Redis, Memcached)

  • Real-time big data analytics

  • High-performance computing (HPC) with large data sets

 Examples

Family Example Types Highlights
r-series r6g, r6i, r7i High memory per vCPU, ideal for databases.
x-series x2idn, x2iedn Very large memory footprint, suitable for SAP HANA.
u-series u-6tb1.metal Ultra-high memory for enterprise-grade workloads.

 Example Use Case

A financial analytics company runs SAP HANA on x2idn.32xlarge for in-memory data processing.

4. Storage Optimized Instances

Description

Designed for high, sequential read/write access to large datasets on local storage (NVMe/SSD).

Best For

  • NoSQL databases (Cassandra, MongoDB)

  • Data warehousing

  • Big data processing

  • Log or file storage

 Examples

Family Example Types Highlights
i-series i3, i4i High IOPS SSD storage.
d-series d3, d2 High HDD storage capacity.
h-series h1 Optimized for throughput-heavy workloads.

 Example Use Case

An analytics platform processes Petabytes of log data using i4i.4xlarge instances for high disk I/O.

5. Accelerated Computing Instances (GPU & FPGA)

 Description

Provide GPU or FPGA hardware acceleration for demanding compute workloads.

 Best For

  • Machine Learning & Deep Learning

  • Video rendering & encoding

  • Scientific simulations

  • Game streaming

Examples

Family Example Types Highlights
p-series p4, p5 GPU-optimized for deep learning.
g-series g5, g6 Graphics and ML inference workloads.
inf-series inf1, inf2 AWS Inferentia chips for ML inference.
f-series f1 Custom FPGA acceleration.

🚀 Example Use Case

A research lab trains a Computer Vision model on p4d.24xlarge GPUs.

6. High Performance Computing (HPC) Instances

Description

Built for massively parallel, high-speed computing tasks that require strong CPU/GPU interconnects.

 Best For

  • Computational fluid dynamics (CFD)

  • Genomics

  • Weather forecasting

  • Seismic modeling

 Examples

Family Example Types Highlights
hpc-series hpc6a, hpc7g High networking bandwidth (up to 200 Gbps).

Example Use Case

A research institute runs climate modeling simulations using hpc7g instances.

7. Arm-based Instances (Graviton)

 Description

Use AWS Graviton processors (ARM-based) — lower cost and power-efficient.

 Best For

  • Web servers, containers, and microservices

  • Open-source databases

  • Cloud-native applications

 Examples

Family Example Types Highlights
t4g, m7g, c7g, r7g Graviton3-based Up to 40% better price-performance ratio.

Example Use Case

A SaaS startup migrates microservices to t4g.medium instances to save 30% on costs.

Quick Reference Table

Category Instance Families Key Feature Example Use Case
General Purpose t, m Balanced Web servers, dev/test
Compute Optimized c High CPU Gaming, APIs, HPC
Memory Optimized r, x, u High RAM Databases, analytics
Storage Optimized i, d, h Fast I/O Data lakes, NoSQL
Accelerated (GPU) p, g, inf, f GPU/FPGA ML, rendering
High Performance Computing hpc Parallel compute Scientific modeling
Graviton (ARM) t4g, m7g, r7g Cost-efficient Containers, web apps

How to Choose the Right Instance Type

  1. Identify workload type (compute, memory, storage, GPU).

  2. Decide performance needs (throughput, latency).

  3. Estimate budget (On-Demand vs Savings Plans).

  4. Start small and scale using Auto Scaling Groups.

  5. Benchmark performance before production deployment.

#!/bin/bash

# Update packages

yum update -y

 

# Install Apache

yum install httpd -y

 

# Start Apache service

systemctl start httpd

systemctl enable httpd

 

# Fetch instance details

INSTANCE_ID=$(curl -s http://169.254.169.254/latest/meta-data/instance-id)

PUBLIC_IP=$(curl -s http://169.254.169.254/latest/meta-data/public-ipv4)

AVAIL_ZONE=$(curl -s http://169.254.169.254/latest/meta-data/placement/availability-zone)

Leave a Reply

Your email address will not be published. Required fields are marked *

Scroll to Top