AWS S3 Interview Questions

Core Concepts

Q: What is S3? Explain storage classes.

S3 is AWS's object storage — unlimited capacity, 99.999999999% (11 nines) durability.

Storage Class	Use Case	Retrieval	Cost

|---|---|---|---|

S3 Standard	Frequently accessed	Instant	Highest
S3 Intelligent-Tiering	Unknown/changing access	Instant	Auto-optimises
S3 Standard-IA	Infrequent access	Instant	Lower storage, retrieval fee
S3 One Zone-IA	Infrequent, single AZ	Instant	20% cheaper than IA
S3 Glacier Instant	Archive, accessed quarterly	Instant	Cheap
S3 Glacier Flexible	Archive, retrieved in hours	1-12 hrs (needs verification — exact retrieval-tier timing changes over time)	Very cheap
S3 Glacier Deep Archive	Long-term archive	12-48 hrs (needs verification — exact retrieval-tier timing changes over time)	Cheapest

Lifecycle policies automate transition between classes.

---

Q: S3 security — how do you secure a bucket?

1.Block Public Access (enabled by default) — prevents accidental public exposure

2.Bucket Policy — resource-based policy attached to bucket

3.IAM Policy — identity-based, who can access

4.ACLs — legacy, avoid for new buckets

5.S3 Object Ownership — enforce bucket owner owns all objects

6.Encryption: SSE-S3 (AWS managed), SSE-KMS (customer managed keys), SSE-C (customer provides key), client-side

json

// Bucket policy: enforce HTTPS only
{
  "Effect": "Deny",
  "Principal": "*",
  "Action": "s3:*",
  "Resource": "arn:aws:s3:::my-bucket/*",
  "Condition": {"Bool": {"aws:SecureTransport": "false"}}
}

---

Q: S3 replication — CRR vs SRR?

CRR (Cross-Region Replication): Replicate objects to a bucket in another region. Use for DR, latency, compliance.

SRR (Same-Region Replication): Replicate within same region. Use for log aggregation, test/prod data sync.

Requirements: Versioning must be enabled on both source and destination.

---

Q: S3 performance optimisation.

•Multipart upload: For objects >100MB. Upload in parallel parts. Required for >5GB.

•Transfer Acceleration: Routes through CloudFront edge locations for faster uploads from geographically distant clients.

•Prefix sharding: S3 supports up to 3,500 PUT/COPY/POST/DELETE and 5,500 GET/HEAD requests per second, per prefix, and automatically partitions a bucket across many prefixes as request rates increase — there's no manual step needed for this to kick in. Manually randomizing key prefixes is largely legacy advice from before S3's automatic scaling (announced 2018); it's now only relevant if you need to burst past a single prefix's limits immediately, since S3's auto-partitioning takes some time to react to a sudden spike.

•S3 Select: Query CSV/JSON data in S3 without downloading entire object (SQL-like).

Revision Notes

S3: object storage, 11 nines durability, unlimited capacity
STORAGE CLASSES: Standard → IA → Glacier (cost drops, retrieval time increases)
Lifecycle policies automate transitions
SECURITY: Block Public Access + Bucket Policy + IAM + Encryption
SSE-S3 (default) | SSE-KMS (audit trail) | SSE-C (you manage key)
REPLICATION: CRR (cross-region DR) | SRR (same region log agg) — needs versioning
PERFORMANCE: Multipart (>100MB) | Transfer Accel | Prefix sharding | S3 Select

S3 — Interview Q&A

AWS S3 Interview Questions

Core Concepts

Revision Notes