EC2 Storage – EBS Volumes

  • EBS = Elastic Block Store
  • A network drive you can attach to EC2 instances.
  • Persists data even after the instance is terminated (unlike instance store).
  • Think of it like a network USB stick.

Key Properties

  • AZ-bound: An EBS volume is created in a specific Availability Zone (e.g., us-east-1a) and can only be attached to EC2 instances in the same AZ.
  • One instance at a time (at CCP level): A single EBS volume cannot be attached to multiple EC2s simultaneously.
  • Attach/detach flexibility: Can detach from one EC2 and attach to another in the same AZ.
  • Capacity must be provisioned in advance: Specify storage size (GB) and IOPS (I/O operations per second). You pay for provisioned capacity.
  • Performance can be increased later by resizing or changing volume type.

Free Tier

  • 30 GB of free EBS storage per month (General Purpose SSD or Magnetic).

Network Latency

  • Since EBS communicates over the network, expect slight latency compared to local storage.

Snapshots

  • EBS Snapshots allow moving volumes across AZs (and even regions).

Multiple Volumes

  • An EC2 instance can have multiple EBS volumes attached (like multiple USB sticks).
  • Each EC2 instance usually has its own root volume, but you can add more.

Unattached Volumes

  • EBS volumes can exist unattached (not linked to any EC2) until needed.

Delete on Termination

  • Attribute controlling EBS lifecycle when instance is terminated:
    • Root volume: By default, deleted when EC2 instance terminates.
    • Additional volumes: By default, not deleted.
    • Can manually enable/disable this behavior.
  • Use case: Keep root volume to preserve logs/data after instance termination.

EBS Volumes (Hands-on Demo)

Viewing EBS Volumes

  • Each EC2 instance has a root device (block device, usually 8 GB by default).
  • Can view attached volumes under EC2 → Storage tab.
  • Clicking the volume opens the EBS Volumes console.

Creating Additional Volumes

  • Can create new EBS volumes in the Volumes section.
  • Must specify:
    • Volume type (GP2, GP3, etc.).
    • Size (e.g., 2 GB).
    • Availability Zone (AZ) – must match the instance’s AZ to attach.

Attaching Volumes

  • After creation, new volumes are in Available state.
  • Must be attached manually to an EC2 instance.
  • Once attached, the EC2 instance will show multiple block devices.
  • To use the new volume: requires formatting & mounting (out of scope for CCP).

AZ Boundaries

  • EBS volumes are AZ-specific.
  • If volume is in different AZ, it cannot be attached to the instance.
  • Example: Instance in eu-west-1b cannot attach volume created in eu-west-1a.

Deleting Volumes

  • Unattached volumes can be deleted instantly.
  • Demonstrates cloud flexibility: create, attach, detach, delete quickly.

Delete on Termination Attribute

  • Root volume:
    • By default, Delete on Termination = Yes.
    • This means when the instance is terminated, the root volume is deleted.
  • Additional volumes:
    • By default, Delete on Termination = No.
    • These persist after instance termination unless manually deleted.
  • Behavior can be changed during instance launch (Advanced storage settings) or later.

Demo Recap

  • Instance terminated → root volume deleted.
  • Additional volume remained intact.
  • Confirms lifecycle behavior of Delete on Termination.

EBS Snapshots

  • Snapshot = backup of an EBS volume at a point in time.
  • Can restore a new volume from snapshot even if original volume is deleted.
  • Recommended (not required) to detach/stop instance before snapshot for data consistency.

Why Snapshots?

  • Restore EBS volume to previous state.
  • Copy across Availability Zones (AZs) or Regions.
  • Enables data transfer using AWS global infrastructure.

How It Works

  1. Take snapshot of EBS volume (optional: stop EC2 first).
  2. Snapshot stored in Region.
  3. Use snapshot to create a new EBS volume in same or another AZ.
  4. Attach new volume to EC2 instance.

Features

  1. EBS Snapshot Archive
    • Move snapshots to cheaper archive tier (75% cheaper).
    • Restore time: 24–72 hours.
    • Use case: long-term, low-priority backups.
  2. Recycle Bin for Snapshots
    • Normally, deleted snapshots are gone permanently.
    • With recycle bin: recoverable for 1 day to 1 year.
    • Protects against accidental deletions.

Amazon Machine Image (AMI)

  • AMI = Amazon Machine Image, powers EC2 instances.
  • Represents a customized template of an EC2 instance.
  • Contains:
    • OS (Linux, Windows, etc.)
    • Software packages / monitoring tools
    • Custom configurations

Benefits

  • Faster boot & configuration (software pre-installed).
  • Reusable for launching multiple identical instances.
  • Region-specific, but can be copied across regions.

Types of AMIs

  1. Public AMIs – provided by AWS (e.g., Amazon Linux 2).
  2. Custom AMIs – created and maintained by you.
  3. Marketplace AMIs – created by vendors, may be free or paid.

AMI Lifecycle / Process

  1. Launch & customize an EC2 instance.
  2. Stop instance (for data integrity).
  3. Create AMI (automatically creates EBS snapshots).
  4. Launch new instances from AMI in same or another AZ/Region.

Use Case Example

  • Launch EC2 in us-east-1a, customize → create AMI → launch identical instance in us-east-1b.

EC2 Image Builder

  • Service to automate creation, maintenance, validation, and testing of:
    • EC2 AMIs (virtual machine images)
    • Container images
  • Removes manual effort in customizing and updating images.

How it Works

  1. Builder EC2 instance created automatically.
    • Installs software (Java, CLI tools, firewalls, apps).
    • Applies OS updates and configurations.
  2. AMI is built from the Builder instance.
  3. Test EC2 instance launched from AMI.
    • Runs pre-defined tests (functionality, security, app health).
    • Testing step is optional.
  4. AMI is distributed (multi-region support for global workloads).

Key Features

  • Automation of build → test → distribute.
  • Can run on schedule (weekly, on package updates, or manual).
  • Free service – only pay for:
    • EC2 instances used during build/test.
    • AMI storage and replication across regions.

Exam Tips

  • EC2 Image Builder = automated AMI lifecycle.
  • Builder EC2 (creates AMI) vs Test EC2 (validates AMI).
  • Supports global distribution of AMIs.
  • Free service, but resources (EC2 + storage) are billed.

Here are concise AWS CCP notes from your EC2 Instance Store transcript:

EC2 Instance Store

  • Physical disk storage directly attached to the host server of the EC2 instance.
  • Provides very high I/O performance compared to EBS.
  • Also called ephemeral storage (data does not persist).

Key Characteristics

  • Fastest storage option for EC2 (millions of IOPS possible).
  • Data is lost when:
    • Instance is stopped or terminated.
    • Underlying host hardware fails.
  • Not durable, not persistent.

Use Cases

  • Temporary or transient data:
    • Buffers, caches, scratch space.
    • Data that can be recomputed or retrieved.
  • Not suitable for long-term storage → use EBS instead.

Limitations & Risks

  • Data loss on stop/terminate/failure.
  • Responsibility for backups and replication is on you.

Example

  • I3 instance family → comes with Instance Store volumes.
  • Comparison:
    • Instance Store → up to 3.3 million IOPS.
    • gp2 EBS volume → up to 32,000 IOPS.

Exam Tip

  • If question mentions “very high performance, non-persistent storage” → Answer: EC2 Instance Store.
  • If question mentions “durable, long-term storage” → Answer: EBS.

Amazon Elastic File System (EFS)

  • Managed NFS (Network File System) by AWS.
  • Can be mounted to hundreds of Linux EC2 instances simultaneously.
  • Works across multiple AZs in a region → highly available and scalable.

Characteristics

  • Shared file system → all instances see the same files.
  • Durable and highly available.
  • Expensive (≈ 3x gp2 EBS cost).
  • Pay-per-use → no capacity planning needed, pay only for data stored.

Comparison: EBS vs EFS

  • EBS: one instance, one AZ, bound to AZ, snapshots for migration.
  • EFS: many instances, multiple AZs, shared access, no snapshots needed.

Storage Classes

  1. EFS Standard – default, frequently accessed files.
  2. EFS Infrequent Access (EFS-IA) – up to 92% lower cost for less frequently used files.
    • Uses lifecycle policies to move files automatically (e.g., after 60 days).
    • Transparent to applications (apps access files the same way).

Exam Tips

  • If question = shared, scalable storage across AZs → EFS.
  • If question = Linux only shared file system → EFS.
  • Remember EFS-IA lifecycle policy for cost optimization.

Key phrase to recall: EFS = managed, shared, scalable NFS for Linux EC2, cross-AZ, with Standard & Infrequent Access storage classes.

Here are concise AWS CCP notes from your EC2 storage shared responsibility model transcript:

EC2 Storage – Shared Responsibility

AWS Responsibilities

  • Infrastructure & durability: Ensures data replication across hardware for EBS/EFS.
  • Hardware replacement: Fixes failed disks, handles infrastructure failures.
  • Data privacy: Prevents AWS employees from accessing your data.

Customer Responsibilities

  • Backups & snapshots: Set up backup plans and snapshot procedures.
  • Encryption: Enable encryption for extra security.
  • Data management: Responsible for all data written to the disk.
  • Instance Store usage:
    • Data lost if instance is stopped or terminated.
    • Must plan for backup of ephemeral storage.

✅ Exam takeaway:

  • AWS secures the infra & replication, customers secure data, backups, encryption, and usage decisions.

Amazon FSx

  • Managed third-party high-performance file systems on AWS.
  • Alternative to EFS (Linux NFS) and S3 (object storage).
  • Offers different flavors for specific workloads.

Main Flavors (for exam)

  1. FSx for Windows File Server
    • Fully managed, Windows-native shared file system.
    • Uses SMB protocol and NTFS.
    • Supports Windows EC2 instances and on-premises clients.
    • Integrated with Microsoft Active Directory.
    • Deployed across multiple AZs for reliability.

2. FSx for Lustre

  • Use cases: Machine Learning, Analytics, Video Processing, Financial Modeling.
  • Features:
    • Hundreds of GB/s throughput. - Millions of IOPS. - Sub-millisecond latency.
  • Can connect to AWS compute instances or on-premises.
  • Data can be linked to Amazon S3 buckets for backend storage.

Other flavors (FYI, not main exam focus)

  • FSx for NetApp ONTAP
  • FSx for OpenZFS

Exam Tips

  • If question = Windows-native file sharing, SMB, NTFS, Active Directory → FSx for Windows File Server.
  • If question = High performance computing, ML, analytics, financial modeling → FSx for Lustre.

✅ Key phrase: FSx = managed specialized file systems (Windows for SMB/NTFS, Lustre for HPC/ML).

Summary