AWS EC2

Table of contents

  1. EC2
  2. EC2 Basics
  3. Launching and Managing EC2 Instances
  4. Elastic Block Store (EBS)
  5. Networking
  6. Monitoring and Management
  7. EC2 Pricing Model
    1. On Demand
    2. Reserved
    3. Spot
    4. Dedicated
  8. AWS Placement Groups Cheat Sheet
    1. What is an AWS Placement Group?
    2. Types of AWS Placement Groups:
    3. Key Considerations:
    4. Usage Tips:
    5. AWS CLI Commands:

EC2

  • Elastic Cloud Computing (EC2) is a Cloud Computing Service.
  • Configure your EC2 by choosing your OS, Storage, Memory, Networking Throughput.
  • Launch and SSH into your serve within minutes.
  • EC2 comes in a variety Instance Types specialized for different roles:
    • General purpose balance of compute, memory and networking resources.
    • Compute Optimized ideal for compute bound applications that benefit from high performance processor.
    • Memory Optimized fast performance for workloads that process large data sets in memory.
    • Accelerated Optimized high, sequential read and write access to very large data sets on local storage.
  • Instance Size generally double in price and key feature.
  • Placement Groups let you choose the logical placement of your instances to optimize communication, performance or durability. Placement Groups are free.
  • User Data a script that will be automatically run when launching an EC2 instance.
  • MetaData meta data about the current instance. You access this meta data via local end-point when SSH into the EC2 instance. eg sh curl http://169.254.169.254/latest/meta-data
  • Instance Profile a container for an IAM role that you can use to pass role information to an EC2 instance when the instance starts.

EC2 Basics

  1. EC2 Instance Types:
    • EC2 instances come in various types optimized for different workloads (e.g., t2.micro, m5.large, c5.xlarge).
    • EC2 instance Types
  2. Regions and Availability Zones:
    • Choose the AWS region and availability zone carefully for fault tolerance and proximity to users.
  3. Amazon Machine Images (AMIs):
    • AMIs are pre-configured templates for creating EC2 instances. Choose an AMI that fits your requirements.
  4. Security Groups:
    • Security groups act as virtual firewalls for EC2 instances. Define inbound/outbound rules to control network traffic.
  5. Key Pairs:
    • Use key pairs to securely access EC2 instances. Create and store the private key securely.

Launching and Managing EC2 Instances

  1. EC2 Launch Wizard:
    • AWS provides an EC2 Launch Wizard to simplify instance setup for specific use cases.
  2. Launch Templates:
    • Use launch templates to standardize instance configurations and simplify instance launches.
  3. Auto Scaling:
    • Set up auto scaling groups to automatically adjust the number of instances based on load or a schedule.
  4. Elastic Load Balancing:
    • Use ELB to distribute incoming traffic across multiple EC2 instances for high availability and fault tolerance.
  5. Instance Store vs. EBS:
    • Understand the differences between instance store and Elastic Block Store (EBS) for storage options.
  6. Instance Metadata:
    • EC2 instances can access metadata for configuration details via http://169.254.169.254/latest/meta-data/.

Elastic Block Store (EBS)

  1. EBS Volume Types:
    • EBS offers various volume types such as GP2, IO1, SC1, ST1, and gp3 with different performance characteristics.
  2. Snapshots:
    • Create EBS snapshots for data backups and recovery. Snapshots are incremental and stored in Amazon S3.
  3. EBS Encryption:
    • You can encrypt EBS volumes at rest for enhanced security.
  4. EBS Performance:
    • Optimize EBS performance with appropriate volume type, size, and IOPS (input/output operations per second).

Networking

  1. VPC (Virtual Private Cloud):
    • Isolate your network resources using VPC. Define subnets, route tables, and security groups.
  2. Elastic IP Addresses:
    • Use Elastic IPs for static, public IPv4 addresses. Associate them with EC2 instances to ensure they retain the same IP even if instances are replaced.
  3. VPC Peering:
    • Connect VPCs to enable communication between instances in different VPCs.
  4. Network Interfaces:
    • Attach additional network interfaces to EC2 instances for multi-networking.
  5. Elastic Network Adapter (ENA):
    • ENA is a high-performance network interface for enhanced networking capabilities.

Monitoring and Management

  1. CloudWatch:
    • Use CloudWatch for monitoring EC2 instances, setting up alarms, and collecting performance data.
  2. Instance Metadata and User Data:
    • EC2 instances can retrieve instance metadata and user data for custom configuration during instance startup.
  3. Tagging:
    • Apply tags to EC2 instances for organization and resource management.
  4. EC2 Instance Roles:
    • Assign IAM roles to EC2 instances to grant them permissions to access AWS services.
  5. EC2 Systems Manager:
    • Use Systems Manager for patch management, automation, and maintenance of EC2 instances.

EC2 Pricing Model

On Demand

  • Least commitment
  • Only pay per hour
  • Short term, spiky, unpredictable workloads
  • Cannot be interrupted
  • For first time apps

Reserved

  • Best long-term
  • Up to 75% off
  • Steady state or predictable usage
  • Commit to EC2 over a 1 or 3 years term
  • Can sell unused reserved instances (Reserved Instances Marketplace)
  • Can be shared between multiple account within an organization
  • Standard Up to 75% reduced price, cannot change the RI attributes
  • Convertible Up to 54% reduced pricing, allows you to change RI attributes if greater or equal in value
  • Scheduled You reserve instances for specific time periods. Eg. once a week for a few hours. Savings vary
  • Terms You commit to a 1 year or 3 years contract. The longer the term the greater the savings.
  • Payment Options All Upfront, Partial Upfront, and No Upfront. The greater the upfront the greater the savings.

Spot

  • Biggest savings
  • Up to 90% off
  • Request spare computing capacity
  • Flexible start and end times
  • Can handle interruptions (Sever randomly starting and stopping)
  • For non-critical background jobs
  • Termination Conditions
    • Instances can be terminated by AWS at anytime, you do not get charged for partial hour of usage
    • If you terminate your instance you will be charge for any hour that it ran

Dedicated

  • Most expensive
  • Dedicated servers
  • Can be on-demand or reserved
  • When you need guarantee or isolate hardware (Enterprise requirements)

Certainly! Here’s a cheat sheet for AWS Placement Groups:

AWS Placement Groups Cheat Sheet

What is an AWS Placement Group?

  • An AWS Placement Group is a logical grouping of instances in a way that influences how they are placed on the underlying hardware.
  • It helps control the placement of instances to meet specific requirements for performance, availability, and isolation.

Types of AWS Placement Groups:

  1. Cluster Placement Group:
  • Grouping strategy: Packs instances closely together within a single Availability Zone.
  • Recommended for applications that require low network latency and high network throughput.
  • No instances from different clusters can share the same placement group.
  1. Partition Placement Group:
  • Grouping strategy: Splits instances into logical partitions, each residing on separate underlying hardware.
  • Ideal for distributed and replicated workloads like Hadoop, Cassandra, or Kafka.
  • Each partition has its own set of racks to improve fault tolerance.
  1. Spread Placement Group:
  • Grouping strategy: Spreads instances across underlying hardware, minimizing the risk of concurrent hardware failures.
  • Recommended for applications where instances must be separated to meet compliance and risk requirements.
  • Can span multiple Availability Zones.

Key Considerations:

  • Placement Groups can be used with Amazon EC2 instances.
  • Instances within a placement group must belong to the same VPC.
  • You cannot move an existing instance into a placement group; you must create the instance within the group.
  • Instance types must be compatible for a placement group (e.g., instances with enhanced networking for a Cluster Placement Group).
  • Placement Groups can be used with Auto Scaling groups to maintain the desired placement.

Usage Tips:

  • Cluster Placement Groups are suitable for applications like high-performance computing (HPC) or tightly coupled, data-intensive applications.
  • Partition Placement Groups are best for large distributed applications and databases.
  • Spread Placement Groups offer isolation and high availability, making them suitable for critical applications.
  • Always consider the trade-offs between performance, availability, and fault tolerance when choosing a placement group type.

AWS CLI Commands:

  • Creating a Placement Group:
    aws ec2 create-placement-group --group-name my-cluster-group --strategy cluster
  • Describing Placement Groups:
    aws ec2 describe-placement-groups --group-names my-cluster-group
  • Launching an EC2 Instance into a Placement Group:
    aws ec2 run-instances --image-id ami-12345 --instance-type t2.micro --placement GroupName=my-cluster-group

  • Adding an Existing Instance to a Placement Group (Not Supported): You cannot move an existing instance into a placement group. You must create a new instance within the group.

  • Changing Placement Group for an Auto Scaling Group: You can specify the placement group when creating or updating an Auto Scaling group.