Skip to content
Draft: this topic has not been reviewed yet. Check facts against the linked AWS documentation before relying on them.

Amazon S3 basics — buckets and objects

Solutions Architect – Associate · Domain 3 · Design High-Performing ArchitecturesTask 3.1: Determine high-performing and/or scalable storage solutionsLevel 118 minDraft
Difficulty
Confidence

Amazon Simple Storage Service (Amazon S3) stores files of any kind, in any amount, and lets you read them back over the internet. It is one of the oldest and most widely used AWS services: websites, apps, backups and data lakes all rely on it.

S3 is object storage. Instead of a disk with folders, you have buckets full of objects, each reached by its name through a web API.

Buckets, objects and keys

  • A bucket is a container for objects. You create it in one Region, and its name must be unique across all of AWS, for example example-company-reports.
  • An object is a file plus metadata (information about it, such as its type and when it was uploaded). A single object can be up to 5 TB.
  • Each object has a key, its full name inside the bucket, such as reports/2026/q3.pdf.

There are no real folders: the slashes in a key just look like folders, and the console shows them that way to help you browse. A bucket can hold an unlimited number of objects.

Durability and availability

These two words are easy to mix up:

  • Durability: how unlikely it is that your data is lost. S3 Standard is designed for 99.999999999% (eleven nines) durability, by storing copies across multiple Availability Zones. If you store 10 million objects, you would expect to lose one every 10,000 years on average.
  • Availability: how likely it is that you can reach your data right now. S3 Standard is designed for very high availability, but a brief outage is still possible, and that does not mean data is lost.

Keeping data private and safe

S3 is private by default. Nobody else can read your objects unless you allow it.

  • S3 Block Public Access is turned on for new buckets and stops anyone from accidentally making data public.
  • IAM policies say what your users and roles may do in S3.
  • Bucket policies are attached to a bucket and say who may access it, including other accounts.
  • Encryption: S3 encrypts every new object at rest automatically, and you can choose your own keys in AWS KMS.
  • Versioning keeps earlier versions of an object, so you can recover from accidental overwrites and deletions.

What people use S3 for

Use Why S3 fits
Backups and archives Very durable, cheap storage classes for rarely used data
Static website hosting S3 can serve HTML, CSS, images and JavaScript directly; add CloudFront for speed
Media and user uploads Stores photos and videos for apps, at any scale
Data lakes for analytics Store raw data cheaply, then query it with services like Amazon Athena
Software and log storage Any number of files, reachable by any AWS service

How you pay

S3 pricing has a few parts: the amount stored per month (depending on the storage class), the number of requests (uploads and downloads), and data transferred out to the internet. Uploading data into S3 from the internet is free. See S3 storage classes for how to pay less for rarely used data.

Putting it together: a photo-sharing app

  • Users upload photos; the app stores each one in an S3 bucket with a key like users/123/photo-456.jpg.
  • The bucket stays private; the app gives users short-lived links to view their own photos.
  • Versioning is on, so a deleted photo can be restored.
  • Photos older than a year move automatically to a cheaper storage class with a lifecycle rule.

Common mix-ups

  • S3 vs EBS. S3 is object storage reached over the web by any client. EBS is a disk attached to one EC2 instance.
  • Durability vs availability. Durability is about not losing data; availability is about reaching it now.
  • Bucket names are global, buckets are regional. The name must be unique everywhere, but the data lives in the Region you choose.

Check yourself

What is the largest single object you can store in S3?

5 TB.

A team accidentally overwrote an important file in S3. Which feature would have let them recover the old version?

Versioning.

Are new S3 buckets public or private by default?

Private. S3 Block Public Access is also on for new buckets.

What does eleven nines of durability describe?

How unlikely it is that stored data is lost: 99.999999999%.

Exam tip for Cloud Practitioner

“Store any amount of files, accessed over HTTP” or “static website” or “data lake” points to S3. “A disk attached to one EC2 instance” is EBS, not S3. “Recover overwritten objects” is versioning.

Key takeaways

  • S3 is object storage: buckets hold objects, each found by its key; objects up to 5 TB, unlimited count.
  • Designed for eleven nines of durability by storing data across Availability Zones.
  • Private by default, encrypted at rest, with Block Public Access, policies and versioning to keep data safe.

Cloud Study Lab is an independent study site. It is not affiliated with or endorsed by Amazon Web Services. AWS and the names of AWS certifications are trademarks of Amazon.com, Inc. or its affiliates.