Amazon EC2 Instance Storage

AWS CLF-C02 — Domain 3: Cloud Technology & Services

Amazon EBS (Elastic Block Store)

EBS is a network drive (not a physical drive) that you can attach to your EC2 instances while they run.

Allows your instance to persist data even after termination
Bound to a specific Availability Zone (e.g., us-east-1a)
Can be detached from one instance and attached to another (within the same AZ)
Uses the network to communicate — there may be a bit of latency
Has a provisioned capacity (size in GBs and IOPS) — you are billed for provisioned capacity
Can increase capacity over time
Delete on Termination: by default ON for root volume, OFF for additional volumes
Think of it like a "network USB stick" that you can plug into one computer at a time

Key Point:

EBS volumes are locked to an AZ. To move a volume to another AZ, you must first create a snapshot, then create a new volume from the snapshot in the target AZ.

EBS Volume Types

Type	Name	Max IOPS	Max Throughput	Size	Boot?	Best For
`gp3`	General Purpose SSD	16,000	1,000 MB/s	1 GiB – 16 TiB	Yes	Most workloads, system boot, virtual desktops, dev/test
`gp2`	General Purpose SSD	16,000	250 MB/s	1 GiB – 16 TiB	Yes	Same as gp3 (older generation, IOPS linked to size)
`io1`	Provisioned IOPS SSD	64,000	1,000 MB/s	4 GiB – 16 TiB	Yes	Critical business apps needing sustained IOPS, databases
`io2`	Provisioned IOPS SSD	256,000	4,000 MB/s	4 GiB – 64 TiB	Yes	Highest performance SSD, sub-ms latency, mission-critical
`st1`	Throughput Optimized HDD	500	500 MB/s	125 GiB – 16 TiB	No	Big data, data warehouses, log processing
`sc1`	Cold HDD	250	250 MB/s	125 GiB – 16 TiB	No	Infrequently accessed, lowest cost

Exam Tip:

Only SSD volumes (gp2, gp3, io1, io2) can be used as boot volumes. HDD volumes (st1, sc1) cannot be boot volumes. If a question mentions "highest IOPS" or "database workload needing sustained IOPS" → io1/io2 (Provisioned IOPS).

EBS Snapshots

A point-in-time backup of your EBS volume
Stored in Amazon S3 (managed by AWS, not visible in your S3 console)
Not necessary to detach the volume to take a snapshot, but recommended
Can copy snapshots across AZs or Regions — this is how you migrate EBS volumes
Incremental — only changed blocks are saved after the first snapshot

Snapshot Features

Feature	Description
EBS Snapshot Archive	Move snapshot to an "archive tier" that is 75% cheaper. Takes 24-72 hours to restore.
Recycle Bin	Protect against accidental deletion. Set retention rules (1 day to 1 year). Deleted snapshots go to recycle bin.
Fast Snapshot Restore (FSR)	Force full initialization of snapshot to have no latency on first use. Expensive.

Exam Tip:

"How to move an EBS volume to another AZ?" → Create a snapshot, then restore it in the new AZ. "How to move to another Region?" → Copy the snapshot to the target Region, then create a volume.

EBS Encryption

When you create an encrypted EBS volume, you get:
- Data at rest is encrypted inside the volume
- All data in flight between the instance and the volume is encrypted
- All snapshots are encrypted
- All volumes created from the snapshot are encrypted
Encryption uses AWS KMS keys (AES-256)
Encryption has minimal impact on latency
Handled transparently by EC2 — you don't need to do anything extra

How to Encrypt an Unencrypted EBS Volume

Create a snapshot of the unencrypted volume
Copy the snapshot and enable encryption on the copy
Create a new volume from the encrypted snapshot
Attach the new encrypted volume to the instance

Amazon Machine Image (AMI)

An AMI is a customization of an EC2 instance — it includes the OS, software, configuration, and data
Provides faster boot time because all software is pre-packaged
AMIs are Region-specific (but can be copied across Regions)
You can launch instances from:
- Public AMI: AWS-provided (e.g., Amazon Linux 2)
- Your own AMI: Custom, made and maintained by you
- Marketplace AMI: Made by third-party vendors (may have a cost)

AMI Process

Start an EC2 instance and customize it
Stop the instance (for data integrity)
Build an AMI — this also creates EBS snapshots
Launch new instances from this AMI

Exam Tip:

AMIs are Region-specific. To use an AMI in another Region, you must copy it to that Region first. AMIs speed up deployment because software is pre-installed.

EC2 Instance Store

Hardware disk physically attached to the host computer (not network-attached like EBS)
Provides very high I/O performance (better than EBS)
Ephemeral storage — data is lost when the instance is stopped or terminated
Good for buffer, cache, scratch data, temporary content
Risk of data loss if hardware fails — you are responsible for backup and replication

Key Point:

Instance Store = ephemeral (temporary) + highest performance. If you see "highest disk I/O performance" in an exam question, think Instance Store. If you need data to persist, use EBS.

Amazon EFS (Elastic File System)

Managed NFS (Network File System) that can be mounted on many EC2 instances simultaneously
Works across multiple Availability Zones (unlike EBS which is AZ-locked)
Only works with Linux instances (POSIX file system)
Highly available, scalable, expensive (about 3x the price of gp2)
Pay per use — no capacity planning needed (automatically scales)
Uses NFSv4.1 protocol
Uses Security Groups to control access

EFS Storage Classes

Storage Class	Description	Cost
EFS Standard	Frequently accessed files	Higher per-GB cost
EFS Infrequent Access (EFS-IA)	Lower cost for infrequently accessed files. Files moved automatically via Lifecycle Policies.	Up to 92% lower cost

Exam Tip:

"Shared file storage across multiple AZs for Linux?" → EFS. "Shared file storage for Windows?" → FSx for Windows File Server. EFS does NOT work with Windows.

Storage Comparison

Feature	EBS	EFS	Instance Store
Type	Network block storage	Network file system (NFS)	Physical disk on host
AZ Scope	Single AZ	Multi-AZ	Single instance
Attach to	One instance (io1/io2: multi-attach)	Many instances	One instance
Persistence	Persistent	Persistent	Ephemeral (lost on stop/terminate)
OS Support	Linux & Windows	Linux only	Linux & Windows
Pricing	Provisioned capacity	Pay per use	Included with instance
Performance	Good	Good	Highest I/O

Amazon FSx

Fully managed third-party high-performance file systems on AWS.

FSx for Windows File Server

Fully managed Windows-native shared file system
Built on Windows File Server
Supports SMB protocol and Windows NTFS
Integrated with Microsoft Active Directory
Can be accessed from on-premises infrastructure

FSx for Lustre

Fully managed high-performance parallel file system for Linux
Lustre = Linux + Cluster
Use cases: machine learning, HPC, video processing, financial modeling
Scales to hundreds of GB/s throughput and millions of IOPS
Can read data directly from S3 and write results back to S3

Exam Tip:

"Windows shared file system with Active Directory?" → FSx for Windows File Server. "High-performance computing on Linux?" → FSx for Lustre. "Linux shared file system, multi-AZ?" → EFS.