S3 Buckets and Objects

Buckets and objects are the two fundamental building blocks of Amazon S3. Mastering their behaviour, naming conventions, metadata handling, and upload strategies is critical to designing effective storage solutions.

---

Bucket Naming Rules

S3 bucket names are subject to strict rules because they form part of the URL used to access objects:

Rule	Detail
Globally unique	No two buckets across all AWS accounts can share a name
Length	3 to 63 characters
Characters	Lowercase letters, numbers, and hyphens only
Start/end	Must start and end with a letter or number
No IP format	Cannot resemble an IP address (e.g. 192.168.1.1)
No periods	Avoid periods — they break HTTPS virtual-hosted-style URLs

Good examples: my-app-assets, acme-corp-backups-2024, data-lake-raw

Bad examples: My_Bucket (uppercase and underscores), 192.168.5.4 (IP format), -start-with-hyphen

---

Bucket Configuration

When you create a bucket, you can configure several important settings:

Region

Buckets are created in a specific AWS Region. Choosing the right region affects:

• Latency — pick a region close to your users or compute resources
• Compliance — certain regulations require data to stay in a specific country
• Cost — pricing varies slightly between regions

Versioning

Versioning keeps multiple variants of every object. We will cover this in depth in Lesson 4.

Default Encryption

You can configure server-side encryption so that every object uploaded to the bucket is automatically encrypted. Options include SSE-S3, SSE-KMS, and DSSE-KMS.

Tags

Bucket tags are key-value pairs used for cost allocation and organisation:

aws s3api put-bucket-tagging --bucket my-app-assets --tagging 'TagSet=[{Key=Environment,Value=Production},{Key=Team,Value=Platform}]'

---

Object Keys and Prefixes

S3 has a flat namespace — there is no real directory hierarchy. However, you can use prefixes and delimiters to simulate folders.

Example keys:

logs/2024/01/access.log
logs/2024/01/error.log
images/thumbnail/photo1.jpg
images/original/photo1.jpg

Here, logs/2024/01/ is a prefix. The console displays these as folders, but technically they are just part of the object key.

Listing by prefix

aws s3 ls s3://my-app-assets/logs/2024/01/

This returns only objects whose keys start with logs/2024/01/.

Why Prefixes Matter

Prefixes are important for:

• Organisation — logical grouping of related objects
• IAM policies — restrict access to specific prefixes
• Lifecycle rules — apply rules to objects matching a prefix
• Performance — S3 partitions data by prefix for high throughput

---

Object Metadata

Every S3 object carries metadata — key-value pairs that describe the object. There are two kinds:

System Metadata

Managed by S3 automatically: