EC2

Amazon Elastic Compute Cloud.

1. Overview

Amazon EC2 (Elastic Compute Cloud) is a web service that provides resizable compute capacity in the cloud. It's designed to make web-scale cloud computing easier for developers.

Key features:

• Scalable virtual servers in the cloud

• Complete control of your computing resources

• Reduced time required to obtain and boot new server instances

• Pay only for the capacity you actually use

2. EC2 Instance Types

EC2 offers a wide variety of instance types optimized for different use cases:

• General Purpose (e.g., T3, M5)

• Compute Optimized (e.g., C5)

• Memory Optimized (e.g., R5)

• Storage Optimized (e.g., I3, D2)

• Accelerated Computing (e.g., P3, G4)

Each instance type offers different combinations of CPU, memory, storage, and networking capacity.

3. Amazon Machine Images (AMIs)

An AMI is a template that contains a software configuration (operating system, application server, and applications).

• You can select from a library of AMIs or create your own

• AMIs can be shared across AWS accounts

• AMIs are regional but can be copied across regions

4. EC2 Purchasing Options

• On-Demand Instances: Pay for compute capacity by the hour or second with no long-term commitments

• Reserved Instances: Purchase instances in advance for significant discounts (1 or 3-year terms)

• Spot Instances: Bid on spare Amazon EC2 computing capacity for up to 90% off the On-Demand price

• Dedicated Hosts: Physical EC2 servers dedicated for your use

Feature	On-Demand Instances	Reserved Instances	Spot Instances	Dedicated Hosts
Description	Pay for compute capacity by the second	Purchase instances in advance for significant discounts	Bid on spare Amazon EC2 computing capacity	Physical EC2 servers dedicated for your use
Commitment	No long-term commitment	1 or 3-year terms	No long-term commitment	On-Demand or Reserved (1 or 3-year terms)
Price Discount	No discount	Up to 72% compared to On-Demand	Up to 90% compared to On-Demand	Can be combined with Reserved Instances for additional savings
Recommended Use Case	Short-term, spiky, or unpredictable workloads	Steady-state or predictable usage	Flexible start and end times, fault-tolerant or stateless workloads	Compliance requirements, licensing restrictions, or high-performance computing needs
Instance Control	Full control	Full control	Can be terminated by AWS with 2-minute warning	Full control
Capacity Reservation	No	Yes	No	Yes
Billing Granularity	Per second	Pre-pay options: All Upfront, Partial Upfront, No Upfront	Per second	Per hour
Availability	Always available	Limited to capacity in a specific AZ	Available only when there's spare EC2 capacity	Limited availability
Convertibility	N/A	Standard RIs are not convertible, Convertible RIs can change instance family, OS, tenancy, and payment option	N/A	N/A
Transferability	N/A	Can be transferred between AZs in the same region	N/A	N/A

5. EC2 Networking

• Virtual Private Cloud (VPC): Launch EC2 instances in a logically isolated section of the AWS Cloud

• Security Groups: Act as a virtual firewall for your instances to control inbound and outbound traffic

• Elastic IP Addresses: Static IPv4 addresses designed for dynamic cloud computing

6. EC2 Storage Options

• Amazon EBS (Elastic Block Store): Persistent block storage volumes for use with EC2 instances

• Instance Store: Temporary block-level storage for your instance

• Amazon EFS (Elastic File System): Fully managed file storage service for use with EC2

7. EC2 Auto Scaling

Automatically adjust the number of EC2 instances in your deployment according to conditions you define.

• Maintain application availability

• Scale your Amazon EC2 capacity up or down automatically according to conditions you define

• Automatically register new instances to a load balancer

8. Load Balancing

Elastic Load Balancing automatically distributes incoming application traffic across multiple targets, such as EC2 instances.

Types of load balancers:

• Application Load Balancer (ALB)

• Network Load Balancer (NLB)

• Classic Load Balancer (CLB)

9. EC2 Instance Metadata and User Data

• Instance Metadata: Data about your instance that you can use to configure or manage the running instance. Instance metadata can be accessed usign IMDSv1 (Legacy method, less secure):

curl http://169.254.169.254/latest/meta-data/

Common metadata endpoints:

bashCopy/latest/meta-data/ami-id                  # AMI ID
/latest/meta-data/instance-id             # Instance ID
/latest/meta-data/instance-type           # Instance type
/latest/meta-data/local-ipv4              # Private IP
/latest/meta-data/public-ipv4             # Public IP
/latest/meta-data/iam/security-credentials/# IAM role credentials
/latest/meta-data/placement/availability-zone # AZ information

• User Data: Script that you can provide when launching your instance to perform common automated configuration tasks. To get user data (if configured):

bashCopycurl http://169.254.169.254/latest/user-data

10. EC2 Hibernation

EC2 Hibernation preserves the in-memory RAM on persistent storage (EBS).

• Faster boot-up because it doesn't need to reload the operating system

• Instance RAM must be less than 150 GB

• Available for specific instance families (C3, C4, C5, M3, M5, R3, R4, R5)

• Supported on Windows, Amazon Linux 2 AMI, and Ubuntu

• Instances cannot be hibernated for more than 60 days

• Available for On-Demand and Reserved Instances

11. EC2 Placement Groups

Placement groups influence the placement of a group of interdependent instances to meet the needs of your workload.

Types:

• Cluster Placement Group: Packs instances close together inside an Availability Zone (low network latency, high network throughput for HPC). All EC2 instances are placed within a rack.

• Partition Placement Group: Spreads instances across logical partitions (Multiple EC2 instances; HDFS, HBase, and Cassandra). All EC2 instances are placed within a rack. Each partition represents a rack. If a rack fails (hardware failure), it may affect multiple instances on that rack, but only within that partition. Can spread multiple AZs.

• Spread Placement Group: Strictly places instances on distinct hardware (when you need to separate critical EC2 instances to run on a separated hardware). Each instance is placed in its own distinct rack. Each rack has at most one instance. Can spread multiple AZs.

12. EC2 Enhanced Networking

Enhanced networking uses single root I/O virtualization (SR-IOV) to provide high-performance networking capabilities on supported instance types.

13. EC2 Monitoring

• Basic Monitoring: Metric data available automatically in 5-minute periods at no charge

• Detailed Monitoring: Metric data available in 1-minute periods for an additional cost

Here are the key EC2 metrics available by default through Amazon CloudWatch:

Basic Monitoring (5-minute intervals):

1. CPU:

• CPUUtilization (%)

1. Disk:

• DiskReadOps (Count)

• DiskWriteOps (Count)

• DiskReadBytes (Bytes)

• DiskWriteBytes (Bytes)

1. Network:

• NetworkIn (Bytes)

• NetworkOut (Bytes)

• NetworkPacketsIn (Count)

• NetworkPacketsOut (Count)

1. Status Checks:

• StatusCheckFailed (Count)

• StatusCheckFailed_Instance (Count)

• StatusCheckFailed_System (Count)

Detailed Monitoring (1-minute intervals, additional cost):

• Same metrics as above but with higher frequency

• Must be explicitly enabled

• Additional charges apply

Important Notes:

• Memory utilization is NOT included by default

• Disk metrics only available for instance store volumes

• For EBS volumes, you need to check EBS-specific metrics

• CPU credit metrics available only for T-class instances:

  • CPUCreditUsage

  • CPUCreditBalance

  • CPUSurplusCreditBalance

  • CPUSurplusCreditsCharged

To get memory, disk space, and other system-level metrics, you need to:

1. Install CloudWatch Agent

2. Configure custom metrics collection

3. Pay for custom metrics