AWS Module 7 Notes

Module 7: Storage

Section 1: Core AWS Storage Services Overview

Section 2: Instance Store (Ephemeral Storage)

• Temporary storage physically attached to EC2 instances.

• Data is lost when the instance stops or is terminated; not suitable for persistent storage.

Section 3: Amazon EBS (Elastic Block Store)

• Provides persistent block storage volumes for EC2 instances.

• Auto-replicated within the same Availability Zone (AZ).

• Offers low-latency, high-performance storage.

• Supports snapshots for backups (incremental after baseline).

Types of EBS Volumes

• General Purpose SSD (gp3/gp2): Default choice for most workloads.

• Provisioned IOPS SSD (io2/io1): High-performance for mission-critical apps.

• Throughput-Optimized HDD (st1): Ideal for large, sequential workloads.

• Cold HDD (sc1): Lowest cost, for infrequent access workloads.

EBS Pricing Considerations

• Volumes: Charged by provisioned GB/month.

• IOPS: Charged by provisioned IOPS (if using io1/io2).

• Snapshots: Charged by GB-month.

• Data Transfer: Charged for inter-region transfers.

Section 4: Amazon S3 (Simple Storage Service)

• Scalable object storage with global access via URL.

• Designed for 99.999999999% durability and 99.99% availability.

• Stores objects in buckets, each up to 5 TB.

• Offers multiple storage classes:

  • S3 Standard: For frequently accessed data.

  • Intelligent-Tiering: Automatically shifts objects between access tiers.

  • Standard-IA: Infrequently accessed data.

  • One Zone-IA: Same as IA but in one AZ (cheaper, lower durability).

  • Glacier / Glacier Deep Archive: Long-term archiving with retrieval delays.

Lifecycle Management

• Lifecycle policies can automatically move objects between storage classes (e.g., Standard → IA → Glacier).

S3 Pricing

• Charged by:

  • Storage (GB per month)

  • Number of requests (e.g., GET, PUT, COPY per 1,000 requests)

  • Data transferred out of AWS

Section 5: Amazon EFS (Elastic File System)

• Scalable, elastic file storage for Linux EC2 instances.

• Shared access across multiple instances (NFS protocol).

• Automatically grows/shrinks based on usage.

• Stores data across multiple AZs by default.

• Best for shared web content, big data workloads, or containers.

Section 6: Amazon S3 Glacier & Glacier Deep Archive

• Long-term, low-cost archival storage.

• Glacier: Retrieval in minutes to hours.

• Glacier Deep Archive: Lowest cost, retrieval can take 12+ hours.

Key Concepts

• Archive: Individual data object.

• Vault: Container for multiple archives.

Retrieval Options

• Expedited (1–5 mins): Highest cost

• Standard (3–5 hrs): Default

• Bulk (5–12 hrs): Lowest cost

Use Cases

• Archiving media, compliance records, health data backups

Note: Glacier & Glacier Deep Archive are S3 storage classes, not separate services.

Section 7: S3 vs. Glacier Comparison

Module Summary

This module explained the different AWS storage services:

• EBS: Block storage for EC2

• S3: Scalable object storage for general-purpose use

• EFS: Shared file storage for Linux EC2 instances

• S3 Glacier/Deep Archive: Archival solutions for long-term, infrequent access

You now understand their primary functions, storage types, use cases, performance characteristics, and pricing models. This will help you choose the right solution depending on your app’s needs.