Chapter 41: AWS EBS

AWS EBS

If you think of AWS storage as a supermarket, then S3 is the aisle for photos, videos, backups and website files (object storage), while EBS is the aisle for high-speed hard disks / SSDs that you attach directly to your virtual machines (EC2 instances) — exactly like plugging an external SSD into your laptop.

Let me explain it like we’re sitting together with a whiteboard and a second cup of filter coffee — slow, step-by-step, full of real analogies, Hyderabad startup examples, 2026 reality, pricing, when to choose which volume type, and a clear hands-on feel.

1. What is Amazon EBS? (Simple + Official Definition)

Amazon Elastic Block Store (EBS) is a high-performance, block-level storage service designed to be used with EC2 instances (virtual machines).

• It provides persistent block storage volumes — like virtual hard disks
• You attach an EBS volume to an EC2 instance → the instance sees it as a normal disk (/dev/xvda, /dev/sdf, etc.)
• Data persists even when you stop or terminate the instance
• EBS volumes live inside one Availability Zone (they are AZ-specific)

Official short line (still perfect in 2026): “Amazon EBS provides block level storage volumes for use with EC2 instances. EBS volumes behave like raw, unformatted block devices. You can attach them to instances, and they persist independently from the life of an instance.”

In plain Hyderabad language: Imagine you rent a scooty from Bounce (EC2 instance).

• The scooty itself = EC2 (compute power)
• EBS = the extra hard box you attach to the scooty to carry your laptop, helmet, water bottle, etc. → The box stays safe even if you return the scooty and rent a new one tomorrow → You can move the box to another scooty if needed → You pay rent only for the box size & performance you choose

2. Why Do We Need EBS? (Key Benefits – 2026 Reality)

• Persistence — data survives instance stop/terminate/reboot (unlike instance store)
• High performance — many volume types offer very high IOPS & throughput
• Snapshots — point-in-time backups → create new volumes from snapshots
• Encryption — at-rest encryption by default (KMS-managed)
• Resizable — increase size & performance without downtime (most types)
• Multi-attach (io1/io2) — attach same volume to multiple instances
• Availability — 99.999% SLA (when used with Multi-AZ setups)

3. The Main EBS Volume Types in 2026 (Which One to Choose?)

AWS offers several volume types — choose based on performance, cost, and workload.

2026 quick rule for Hyderabad teams:

• 90% of cases → gp3 (best price/performance, flexible IOPS/throughput)
• Databases with high IOPS needs → io2 or io2 Block Express
• Very large sequential workloads → st1
• Cheap cold storage → sc1 (rare now)

4. Real Hyderabad Example – Full EBS Usage in a Startup

Your food delivery backend startup in Gachibowli:

Components using EBS:

1. EC2 web servers (Node.js API)
   • Root volume: gp3 30 GB (boot + code)
   • Extra volume: gp3 100 GB for temporary file uploads (invoices, photos)
2. RDS PostgreSQL database (orders, users)
   • Primary volume: io2 500 GB, 10,000 provisioned IOPS
   • Multi-AZ → standby in different AZ gets its own EBS volume automatically
3. Redis cache (ElastiCache)
   • Uses EBS gp3 under the hood (you don’t manage it)
4. Daily backup strategy
   • Automated EBS snapshots → keep 7 daily + 4 weekly + 12 monthly
   • Snapshots stored in S3 → very cheap

During IPL final rush:

• Orders spike 8× → database IOPS jump → io2 volume auto-handles (provisioned performance)
• Web servers scale from 4 → 12 instances → each new instance gets fresh gp3 root volume in seconds
• Snapshots let you restore quickly if something goes wrong

Monthly storage cost example (moderate traffic):

• gp3: 2 TB total → ~₹13,600
• io2: 500 GB + 10k IOPS → ~₹6,000–8,000
• Snapshots: ~₹200–500
• Total storage: ₹20,000–25,000 (vs 3–5× more if using only io1 or older types)

5. Quick Hands-On – Create & Attach an EBS Volume

1. EC2 console → Launch instance (t4g.micro)
2. Storage tab → Root volume (gp3, 8 GB default)
3. After launch → Volumes → Create volume (gp3, 20 GB, same AZ as instance)
4. Attach volume → to your instance (/dev/sdf)
5. SSH into instance →
   Bash

   PHP

   0 1 2 3 4 5 6 7 8 9 10

   lsblk               # see xvda (root), xvdf (new) sudo mkfs.ext4 /dev/xvdf sudo mkdir /data sudo mount /dev/xvdf /data df -h               # see it mounted

6. Write files → stop instance → start again → data still there

Summary Table – EBS Cheat Sheet (2026 – India Focus)

Teacher’s final note: EBS is the “hard disk” you attach to your EC2 virtual machines — persistent, reliable, and tunable for performance. In 2026, almost every production workload in Hyderabad that runs on EC2 uses EBS (gp3 for most things, io2 for databases). Don’t confuse it with instance store (temporary) or S3 (object storage) — they solve completely different jobs.

Got it? This is the “where does my database and OS actually live?” lesson.

Next?

• Deep dive: gp3 vs io2 performance & pricing comparison?
• How to create EBS snapshots + automate backups?
• Or EBS Multi-Attach + RAID for high-availability storage?

Tell me — next whiteboard ready! 🚀💽