Chapter 13: Amazon EBS (Elastic Block Store)
13.1 What and why
Mitranno, sooner or later every server runs out of disk space — usually at the worst possible time. This chapter teaches you how to handle storage calmly. Let's first understand what EBS is. Amazon EBS provides block-level storage volumes — virtual hard disks — that you attach to EC2 instances. The root disk of almost every EC2 instance is an EBS volume.
Why EBS?
- Persistent: data survives instance stop/start (unlike instance store, which is lost).
- Flexible: increase size, change type or performance while the volume is in use.
- Durable: automatically replicated within its Availability Zone.
- Backups: point-in-time snapshots stored in Amazon S3 (managed by AWS), copyable across regions.
- Encryption: one tick box (AWS KMS) encrypts data at rest, in transit to the instance and in snapshots.
Availability Zone ap-south-1a
+------------------------------------------------------------+
| EC2 instance |
| ├── /dev/nvme0n1 (root EBS, gp3 8 GiB) → / |
| └── /dev/nvme1n1 (data EBS, gp3 20 GiB) → /data |
+------------------------------------------------------------+
│ snapshot (incremental)
▼
EBS Snapshots (regional, stored in S3) ──► create new volume in ANY AZ of the region
Key rules
- An EBS volume lives in one Availability Zone and can only attach to instances in the same AZ. To move it to another AZ/region, take a snapshot and create a new volume from it.
- Normally one volume attaches to one instance at a time (io1/io2 support Multi-Attach for special clustered workloads).
- You pay for the provisioned size (GiB-month), even if the disk is empty and even while the instance is stopped.