📦 Understanding Amazon EBS (Elastic Block Store) — Simple Guide

1. 💾 What is Amazon EBS?

Amazon EBS (Elastic Block Store) provides persistent block storage for Amazon EC2 instances.

👉 In simple terms:
EBS acts like a virtual hard drive attached to your EC2 instance.

✔ Data persists even if the instance:

• Stops

• Restarts

1.1 🔗 How EBS Works

• Create an EBS volume

• Attach it to an EC2 instance

• Use it like a disk inside the OS

📌 Common uses:

• Operating system (root volume)

• Application data

• Logs and databases

2. 📊 EBS Performance Metrics

Understanding performance is critical for real-world workloads.

(EBS performance is evaluated using four main metrics.)

2.1 IOPS (Input/Output Operations Per Second)

IOPS represents:

• Number of read/write operations a volume can handle per second

Important:

Small I/O operations → higher IOPS demand

• Databases are IOPS-sensitive

📌 Example:

• 1,000 IOPS = 1,000 operations/sec

✔ Important for:

• Databases

• Transaction systems

2.2 Block Size

• Amount of data per I/O operation

Common sizes:

• 4 KB

• 8 KB

• 16 KB

📌 Impact:

• Smaller block size → more IOPS

• Larger block size → fewer IOPS, higher throughput

2.3 Latency

Latency is:

• Time taken to complete one I/O operation

Measured in:

• milliseconds (ms)

Lower latency = faster response Critical for:

• Databases

• Transaction systems

• Real-time applications

2.4 Throughput

Throughput is:

• Rate at which data is transferred

Measured in:

• MB/s

• GB/s

Important for:

• Large file processing

• Data analytics

• Streaming workloads

2.5 Relationship Between IOPS, Throughput, and Block Size

A simplified formula:

IOPS = (Throughput × 1024) / Block Size (KB)

Meaning:

• Increasing block size increases throughput but reduces IOPS

• Workload type decides which metric matters more

3. 📦 EBS Volume Types (Overview)

Three main categories:

• SSD-backed volumes → low latency, random I/O

• HDD-backed volumes → high throughput

• Magnetic volumes → legacy

4. Root Volume vs Additional Volumes

Root Volume

• Contains the Operating System

• Mandatory for launching an EC2 instance

• Without root volume → instance cannot boot

Supported root volume types:

• gp2

• gp3

• io1

• io2

Additional (Data) Volumes

• Used for application data, logs, databases

• Can be attached/detached independently

Supported additional volume types:

• gp2, gp3

• io1, io2

• st1, sc1

• standard (magnetic)

5. SSD-Based EBS Volumes

SSD volumes are optimized for low latency and random I/O.

5.1 General Purpose SSD – gp2 & gp3

These are the most commonly used EBS volumes.

gp2

• Balanced performance

• Performance scales with size

• Suitable for:

  • Boot volumes

  • Web servers

  • Dev/Test workloads

Size range:

• 1 GB to 16 TB

gp3 (Recommended)

• Decouples storage size from performance

• You can independently configure:

  • IOPS

  • Throughput

Size range:

• 1 GB to 64 TiB

Why gp3 is better:

• More predictable performance

• Lower cost

• Better control

5.2 Provisioned IOPS SSD (io1 & io2)

Used when performance must be guaranteed.

Used when exact IOPS guarantees are required.

Suitable for:

• Databases

• High-transaction systems

• Mission-critical workloads

io1

• Provision specific IOPS

• Good for consistent high performance

• Size range: 4 GB to 16 TB

io2

• Higher durability

• Higher IOPS capability

• Designed for critical production systems

Key characteristics:

• Minimum IOPS: 100

• Maximum IOPS: up to 100,000

• Size range: 4 GB to 64 TiB

6. 🧱 HDD-Based Volumes

HDD volumes are optimized for high throughput, not random IOPS.

6.1 Throughput Optimized HDD – st1

Best suited for:

• Big data workloads

• Log processing

• Data warehousing

• Streaming large files

Characteristics:

• High MB/s throughput

• Not suitable for boot volumes

• Lower cost than SSD

Size range:

• 125 GB to 16 TB

6.2 Cold HDD – sc1

Designed for:

• Infrequently accessed data

• Cold storage

• Cost-sensitive workloads

Characteristics:

• Lowest performance

• Lowest cost

• Sequential access only

Size range:

• 125 GB to 16 TB

7. Magnetic (Standard) Volumes

• Legacy storage option.

  Characteristics:

  • Lowest cost

  • Lowest performance

  • Rarely used today

  Size range:

  • 1 GB to 1 TB

  Used only for:

  • Legacy systems

  • Very low-cost requirements

8. 🌍 Availability Zone Limitation (Critical)

⚠️ EBS volumes are Availability Zone–specific

👉 Must match:

• Region

• Availability Zone

📌 Example:

• EC2 in us-east-1a

• EBS must also be in us-east-1a

8.1 🔄 Cross-AZ Migration (Using Snapshots)

EBS volumes cannot be moved directly across Availability Zones. To transfer an EBS volume from one zone to another, create snapshots.

✔ Solution: Snapshots

• Stored internally in Amazon S3

Migration Flow:

1. Create snapshot

2. Copy/use snapshot in another AZ

3. Create new volume

4. Attach to EC2

9. Linux Commands to Work with EBS Volumes

9.1 Detect Attached Volumes

lsblk

Shows:

• All block devices detected by OS

• Even unmounted volumes

Example volume names:

• /dev/nvme0n1 → root volume

• /dev/nvme1n1 → additional volume

9.2 View Mounted File Systems

df -Th

Shows:

• Mounted volumes only

• File system type

• Human-readable sizes

9.3 Check File System Presence

file -s /dev/nvme1n1

Output meaning:

• data → no file system

• XFS / ext4 → file system exists

10. Creating File System and Mounting (Linux)

Create XFS File System

mkfs -t xfs /dev/nvme1n1

Mount the Volume

mount /dev/nvme1n1 /home/ec2-user/newvol

Temporary mount:

• Stored in /etc/mtab

• Lost after reboot

Permanent Mount Configuration

1. Copy mount entry

2. Add it to /etc/fstab

Example:

mount /dev/nvme1n1 /home/ec2-user/newvol
tail -1 /etc/matb
sudo vi /etc/fstab

Activate:

mount -a

11. Expanding an Existing EBS Volume

Step 1: Increase Volume Size (AWS Console)

• Modify volume

• Increase size (cannot decrease)

Step 2: Expand File System (Linux – XFS)

xfs_growfs -d /home/ec2-user/newvol

This updates file system to use new space.

12. Expanding Root Volume (Linux)

Step 1: Increase root volume size in AWS Console

Step 2: Identify Root Partition

Usually:

• Volume: /dev/nvme0n1

• Partition: 1

Step 3: Grow Partition

growpart /dev/nvme0n1 1

Step 4: Reboot Instance

reboot

File system expands after reboot.

13. ⚠️ Important Limitation

❌ AWS does NOT support reducing EBS volume size

If you need smaller size:

• Create new volume

• Copy data

• Replace old volume

14. 🔄 EBS vs Instance Store

Feature	EBS	Instance Store
Persistence	✅ Yes	❌ No
Stop/Start Safe	✅ Yes	❌ No
Type	Network	Physical
Use Case	Durable storage	Temporary data

15. ✅ Best Practices (Real-World)

• Use gp3 by default

• Use io2 for databases

• Avoid st1/sc1 for DBs

• Separate root and data volumes

• Take snapshots before changes

• Monitor via Amazon CloudWatch

📌 Final Takeaway

Amazon EBS is a core storage service in AWS, providing:

• Persistence

• Scalability

• Performance flexibility

But remember:

👉 It is tightly coupled to the Availability Zone, which impacts architecture design.