☁️

Understanding Virtualization and AWS Compute

Mar 17, 2025

Lecture Notes: Virtualization and Compute on AWS

Introduction to Virtualization

  • Virtualization is a technology that allows multiple operating systems to run on a single physical machine.
  • Hypervisors are software that enables virtualization.
  • Types of hypervisors and their use cases were discussed.

Compute on AWS

  • EC2 Instances: AWS's virtual servers.
  • EC2 Architecture: Overview of how AWS virtual machines operate.
  • Dedicated vs. Shared Hosts: Different hosting environments based on needs.

EC2 Instance Types

  • Instance Types: Crucial for selecting appropriate resources (compute, storage, RAM) for application needs.
  • AWS Instance Categories: Five main categories based on resource needs.
    • General Purpose: Balanced resources, ideal for uncertain application needs or testing.
    • Compute Optimized: High compute needs, suitable for gaming and machine learning applications.
    • Memory Optimized: High memory needs, good for in-memory caching applications.
    • Accelerated Computing: Includes GPUs, ideal for graphics-intensive applications like Netflix.
    • Storage Optimized: High-speed local storage, suitable for applications requiring fast data access.

Instance Selection

  • Instance Family and Generation: Naming conventions (e.g., t2.micro) indicate family and generation.
  • Resource Allocation: Based on family and generation, affects compute, storage, etc.

Instance Billing Models

  • On-Demand Pricing: Pay-as-you-go model, most expensive, used for non-downtime critical applications.
  • Spot Pricing: Cheaper, up to 90% savings, but instances can be reclaimed by AWS during high demand.
  • Reserved Pricing: Up to 75% savings with commitment (1-3 years), depends on upfront payment.
  • Savings Plans: Flexible, covers a range of compute resources, similar to reserved pricing.

Compute Storage

  • Types of Storage: Block, file, and object storage.
  • Block Storage: Data chopped into blocks, each with a unique ID, stored in volumes.
    • EBS (Elastic Block Store): Persistent block storage, attachable and detachable.
    • Instance Store: Non-persistent, temporary storage directly attached to hardware.

Performance Metrics

  • IOPS (Input/Output Operations Per Second): Measures read/write operations.
  • Throughput: Measures data transfer speed.

Disaster Recovery

  • Snapshots: EBS feature for disaster recovery, can be transferred across regions.
  • Incremental Snapshots: Only changes since the last snapshot are saved.
  • Data Lifecycle Manager: Automates snapshot scheduling for regular backups.

Practical Application

  • Hands-on with EC2: Creation of an EC2 instance using Windows, setting up user data for installation.
  • Migration and Backup: Simulated backup strategy using EBS snapshots and AMIs.
  • Remote Desktop Protocol (RDP): Used for connecting to Windows instances.

Troubleshooting and Best Practices

  • Importance of knowing regions for managing and finding resources.
  • Using tags correctly to manage and identify AWS resources.

These notes summarize a lecture on virtualization and compute services on AWS, focusing on EC2 instance types, billing models, storage options, and disaster recovery strategies.

Full transcript