SAA-C03 - Design High-Performing Architectures (24% of the exam) - Section 3.1

Determine high-performing object and file storage using Amazon S3, Amazon EFS and Amazon FSx.

Compare Amazon S3 object storage, Amazon EFS network file storage, and Amazon FSx managed file systems and identify which suits each workload by access pattern, throughput requirements, and protocol. Recognise that S3 scales request rates automatically per prefix, and that EFS and FSx for Lustre are appropriate when multiple compute instances need shared POSIX file access at low latency.

Amazon S3Amazon EFSAmazon FSxS3 request rates

Practice question for this objective

Free sampleDesign High-Performing Architecturesmedium

A company is choosing a fully managed AWS file storage service for two separate workloads. Workload one is a Linux analytics fleet that needs a shared POSIX file system mountable over NFS by thousands of EC2 instances at once with elastic capacity. Workload two is a Windows application that needs an SMB share supporting NTFS access control lists and joined to the existing Microsoft Active Directory domain. The team wants each workload on the service purpose-built for it with the least operational overhead. Which two service selections are correct? Select TWO.

AUse Amazon EFS for the Linux analytics fleet because it provides an elastic NFS file system that many EC2 instances can mount concurrently. Correct
BUse Amazon S3 with a mounted file gateway for the Linux analytics fleet so the instances share files through object storage.
CUse Amazon FSx for Windows File Server for the Windows application because it offers SMB shares with NTFS ACLs and Active Directory support. Correct
DUse Amazon EFS for the Windows application by enabling its SMB protocol option and joining the share to Active Directory.
EUse Amazon FSx for Lustre for the Windows application because it is the managed Windows file service with Active Directory support.

Match shared file storage to protocol needs by selecting Amazon EFS for elastic Linux NFS access and Amazon FSx for Windows File Server for SMB with Active Directory. Amazon EFS provides an elastic NFS file system for concurrent POSIX access from many Linux instances, while Amazon FSx for Windows File Server provides native SMB with NTFS ACLs and Active Directory integration; each is purpose-built for its protocol, so pairing them by workload meets both requirements with minimal operational effort.

Why A is correct: Amazon EFS is a managed elastic NFS file system designed for concurrent POSIX access from large Linux fleets, matching workload one exactly.

Why B is wrong: S3 is object storage and a file gateway adds latency and complexity; it is not the purpose-built shared POSIX file system this NFS workload needs.

Why C is correct: FSx for Windows File Server delivers fully managed SMB storage with NTFS permissions and Active Directory support, meeting workload two precisely.

Why D is wrong: Amazon EFS speaks NFS only and has no SMB or NTFS ACL support, so it cannot serve the Windows Active Directory share this workload requires.

Why E is wrong: FSx for Lustre is a high-performance Linux file system for HPC, not a Windows SMB service, so it does not fit the Active Directory requirement.

See more SAA-C03 practice questions, answers explained.

Exam traps in Design High-Performing Architectures

Answers that look right on this material and are not. Each one is a distractor from a different question in the SAA-C03 bank for this domain.

Enable S3 Versioning on the bucket so each request targets a distinct object version and the additional version metadata spreads the load.
Why it is wrong: Versioning keeps prior copies of an object for recovery, but it does not change how S3 partitions request capacity by prefix, so the throughput ceiling on the hot prefix is unchanged.
Attach a separate Amazon EBS volume to each instance and run a synchronisation agent that replicates uploaded files between the volumes on a schedule.
Why it is wrong: EBS volumes are tied to one instance and one Availability Zone, so scheduled replication is fragile, lags real-time writes, and does not give all instances a single consistent view of the content.
Keep Bursting Throughput but store more data so the larger size raises the baseline rate and grows the pool of burst credits the bursts draw down.
Why it is wrong: Padding data to lift the baseline is wasteful and still credit-limited, so sustained bursts exhaust credits and throttle exactly as before.

Examworthy is not affiliated with or endorsed by Amazon Web Services. Original, blueprint-aligned practice material only.