Chapter Two

Storage

Object, block, and file storage are three different models, not three flavors of the same thing. Choosing the wrong one is one of the most common and most expensive AWS mistakes.

7 services

Core Terminology

A few terms recur across this chapter. Defining them here keeps the service sections focused.

Durability
The probability data survives over time. S3 Standard is designed for eleven nines (99.999999999%) — losing an object is statistically negligible.
Availability
The probability data is reachable right now. Distinct from durability: data can be safe but temporarily unreachable.
Object vs Block vs File
Object (S3) is key-and-blob over HTTP; block (EBS) is a raw disk for one instance; file (EFS/FSx) is a shared file system many hosts mount.
Storage Class
A pricing-and-access tier within S3 — Standard, Infrequent Access, Glacier — trading retrieval speed and cost.
Snapshot
A point-in-time, incremental copy of an EBS volume stored in S3, used for backup and cloning.
IOPS / Throughput
Two distinct performance dimensions: operations per second versus megabytes per second. Different workloads bottleneck on different ones.

Services in This Chapter