Amazon AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 66

Question 650

A large EC2 block device stores multi-terabyte sequential log files and requires high streaming throughput but not high random IOPS or bootability. Which EBS tier is economical?

Put a random-write transactional database root filesystem on EBS st1 and assume its sequential throughput design gives premium random IOPS.
Evaluate EBS st1 throughput-optimized HDD for a supported sequential workload instead of premium io2.
Use EBS gp2 and expand the volume solely to raise baseline IOPS despite having a large sequential-throughput workload.
Use EBS sc1 cold HDD as a drop-in guaranteed highest-throughput class without checking the throughput bound of the large streaming job.

Explanation

Correct Answer

st1 emphasizes sustained sequential MB/s throughput on large attached data volumes, not low-latency random transactions.

Incorrect Answers

Throughput HDD is for large sequential I/O and not the suitable boot or high-random-IOPS database volume type.

gp2 is SSD with performance/cost coupling that may be less economical than a supported st1 sequential stream; evaluate the throughput profile.

sc1 is a cold HDD tier with lower throughput and access assumptions than st1; a sustained large sequential workload may require st1 throughput-optimized HDD.

Question 651

A backup vault contains thousands of old EBS snapshots. Owners want to lower snapshot storage while keeping the required restoration points. What is the correct costing approach?

Inventory protected snapshot chains and retention requirements, then expire unneeded snapshots under a verified backup policy.
Delete every available EBS snapshot without checking dependent recovery points or the block retention needed by later snapshots.
Assume deleting the oldest snapshot also deletes every block shared by later restore points.
Copy the same snapshots to another Region and expect the source snapshot storage charge to disappear.

Explanation

Correct Answer

Snapshots store incremental changed blocks; deleting one snapshot need not free its apparent full logical volume size.

Incorrect Answers

EBS snapshots are incremental and recovery dependency/retention requirements must be checked before removing any snapshot.

Snapshot block retention is reference-dependent; blocks required to restore other snapshots remain stored even when an earlier snapshot is removed.

Cross-Region snapshot copies add independent backup storage unless the original retained copies are deliberately reviewed and removed.

Question 652

A sensor fleet creates hundreds of millions of 1-KiB S3 objects daily; stored gigabytes remain modest, but request charges dominate. Which storage design change is worth testing?

Batch small records into larger appropriately partitioned S3 objects while preserving per-record retrieval requirements.
Switch tiny S3 objects to a cheaper storage class while preserving the same hundreds of millions of PUT requests and claiming API charges vanish.
Place every 1-KiB record in S3 Intelligent-Tiering and assume the storage class aggregates the records into fewer billable PUT requests without an application format change.
Run S3 Batch Operations over the tiny existing objects and assume it merges millions of independent object keys into larger records without an explicit transform.

Explanation

Correct Answer

Reducing the number of PUT operations and object metadata overhead can lower request cost for tiny-object workloads.

Incorrect Answers

Object storage rate does not eliminate per-request write charges; fewer suitably batched objects can lower the large request count.

A storage class cannot batch logical records or reduce how many PUTs the application sends. Record aggregation must occur before S3 writes, with read and reprocessing constraints handled explicitly.

Batch Operations can perform supported per-object actions but does not itself aggregate hundreds of millions of micro-objects into application-defined records.

Question 653

An analytics application replicates 20 TB of S3 data from Virginia to Frankfurt daily, including unchanged objects in one design. Which cost model is critical?

Add cross-Region S3 replication indiscriminately and assume destination storage, destination PUT and inter-Region data transfer are free.
Run the same batch S3 copy operation on the already-replicated 20-TB dataset every day without using change detection, assuming unchanged data causes no additional transfer or destination write requests.
Route cross-Region copy traffic through a VPC peering connection and assume private network routing waives S3 replication transfer pricing.
Budget and minimize cross-Region replication transfer/storage and destination-request charges according to recovery need.

Explanation

Correct Answer

Cross-Region Replication creates destination copies and network transfer costs; replication should reflect actual protection objectives.

Incorrect Answers

Replica copies occupy destination storage and may incur transfer and replication-related charges in addition to the source data.

Explicit repeated copies can still generate data movement and destination requests. Use appropriate replication/incremental selection instead of repeatedly copying unchanged objects.

Network topology does not negate S3 cross-Region transfer and destination storage costs for replicated objects.

Question 654

A research foundation exposes a public S3 dataset to many external AWS accounts and wants supported requesters to pay eligible request and transfer costs. Which feature?

Enable Requester Pays but allow anonymous object GET requests without any authenticated requester billing context.
Evaluate S3 Requester Pays with appropriate requester authentication/headers and chargeback expectations.
Enable S3 Transfer Acceleration and assume the service transfers every external customer’s GET and retrieval fees onto their separate AWS account.
Enable bucket versioning and assume each external downloader automatically pays the data owner’s S3 charges.

Explanation

Correct Answer

Requester Pays can shift supported requester-billed retrieval and data transfer charges while the bucket owner retains storage costs.

Incorrect Answers

Requester Pays requests require a supported requester/billing acknowledgment; anonymous or incompatible requests cannot be charged as desired.

Transfer Acceleration concerns network path and transfer pricing; it does not implement requester billing acknowledgements for a Requester Pays bucket.

Versioning governs object history and storage, not which authenticated account is billed under S3 Requester Pays.

Question 655

A new AWS Enterprise Support customer has reserved a Data Transfer Terminal appointment to upload 80 TB from customer-owned portable storage. The team expects to arrive with bare disks and no compatible host or network adapters, relying on AWS staff to operate the media. What preparation is required?

Arrive without a configured laptop, storage controller or network interface because AWS takes possession of the disks and performs a managed Snowball-style transfer for every reservation.
Skip hardware compatibility checks because a confirmed Enterprise Support plan guarantees that any disk and any 1-GbE adapter connects directly to the facility optical network.
Bring supported host and storage equipment, confirm compatible network/optical connectivity and reservation/access requirements, and operate the data transfer through the facility's high-bandwidth connection.
Order a newly shipped Snowball Edge device instead of operating customer media at the terminal, because AWS requires new Enterprise Support customers to use Snowball hardware.

Explanation

Correct Answer

AWS documents reservations, identity/access logistics and compatible customer-operated transfer equipment. Confirming optical transceivers, host connection and destination authentication before arriving avoids consuming a reserved session without a viable transfer path.

Incorrect Answers

Data Transfer Terminal is a physical access facility where eligible customers bring and operate their own devices. AWS does not automatically take bare disks and perform a Snowball shipping workflow for every reservation.

Eligibility to reserve a facility does not guarantee hardware compatibility. The terminal's fiber connection and device preparation requirements must be met with appropriately equipped host hardware and networking.

AWS Snowball Edge is no longer available for new customers. The terminal alternative works with customer-provided equipment and a supported reservation, not a newly shipped Snow Family device.

Question 656

An EFS file system has months-old backups accessed only for quarterly audits and recent hot working files. What lifecycle capability can lower cold-file storage expense?

Transition hot frequently read EFS files to IA without measuring per-access fees, assuming IA reads are always free.
Configure suitable EFS lifecycle transitions to supported Infrequent Access and optionally archive storage classes based on access age.
Move EFS file paths directly into S3 Glacier without migrating file data or changing the filesystem protocol, expecting POSIX opens to work unchanged.
Enable EFS throughput provisioning at a higher baseline and expect the change to lower cold-file GiB-month storage rates.

Explanation

Correct Answer

EFS lifecycle can move eligible infrequently touched files into lower-cost storage while retaining file access semantics.

Incorrect Answers

EFS IA storage savings can be offset by metered read access; lifecycle selection depends on real access patterns.

S3 Glacier is object archive storage, not an automatic EFS file lifecycle tier; EFS offers its own supported IA/Archive classes.

Throughput-mode charges differ from EFS storage-class charges; lifecycle movement addresses cold file storage cost.

Question 657

A new EFS file system has highly variable throughput demand and the team wants to avoid reserving a steady provisioned rate. Which mode should be compared economically?

Use EFS Elastic throughput for a stable, continuously high bandwidth requirement and assume it always costs less than measured provisioned throughput.
Use EFS Bursting throughput but assume every new file system has unlimited burst credits regardless of storage and demand.
Provision maximum EFS throughput for every hour and assume idle periods incur no committed performance charges.
Evaluate EFS Elastic throughput against actual read/write throughput and provisioned/bursting alternatives.

Explanation

Correct Answer

Elastic throughput scales used capacity independently of file-system size, charging for measured operations instead of fixed provisioned throughput.

Incorrect Answers

Elastic throughput has usage-based economics; persistent predictable high traffic can favor provisioned throughput after comparison.

Bursting performance has capacity/credit considerations; Elastic throughput is a measured alternative for variable access.

Provisioned throughput may incur cost for the fixed rate despite variable low periods; compare it with Elastic usage billing.

Question 658

A Windows file workload requires SMB shares with deduplication on repetitive large files and supported managed backups. Which managed filesystem consideration improves cost?

Choose a generic EBS volume mounted on one Windows EC2 server and assume it provides a managed HA SMB file service with deduplication.
Use EFS with NFS-only access and assume it exposes a managed native Windows SMB file share and Windows deduplication controls.
Use FSx for Windows but leave data deduplication disabled, assuming that file-service encryption automatically compresses repeated content.
Evaluate Amazon FSx for Windows File Server with supported storage capacity, deduplication and throughput configuration.

Explanation

Correct Answer

A fit-for-purpose managed SMB filesystem can reduce redundant storage while meeting Windows features and throughput needs.

Incorrect Answers

EBS is a block volume; managed Windows SMB availability and file-level deduplication are FSx for Windows capabilities.

EFS uses NFS, not native Windows SMB shares and file-server deduplication; FSx for Windows is the managed fit.

Encryption protects confidentiality but is not the Windows deduplication feature needed to reduce redundant file storage.

Question 659

A machine-learning service in Region A repeatedly downloads a large dataset from an S3 bucket in Region B. The application needs no cross-Region copy for compliance. What cost/performance tradeoff?

Change the S3 bucket from Standard to Standard-IA while continuing the identical large cross-Region download path each hour.
Add an S3 gateway endpoint in the compute VPC in Region A but continue requesting the remote Region B bucket through the same cross-Region data path, assuming the endpoint moves the data locally.
Enable S3 Intelligent-Tiering for the source and assume the class automatically moves the dataset into the EC2 instance’s Region.
Evaluate placing computation and dataset in the same AWS Region or maintaining a justified local replica.

Explanation

Correct Answer

Repeated inter-Region object access can add transfer cost and latency; locality can remove unnecessary repeated cross-Region reads.

Incorrect Answers

Storage class affects at-rest and retrieval economics; recurring inter-Region bytes remain charged and may need data locality changes.

A gateway endpoint provides supported VPC connectivity but does not relocate an S3 bucket to the compute Region. The Region B object read can still carry cross-Region transfer and latency costs.

Intelligent-Tiering changes access tier, not geographic Region or data locality; repeated cross-Region reads remain.

Question 660

Operations must attach revised compliance tags to millions of existing S3 objects and retain an auditable completion report. Which managed bulk action?

Run an S3 Batch Operations job with an empty manifest and assume it discovers every noncompliant object automatically.
Use S3 Storage Lens to identify objects but assume it writes a new compliance tag onto every old object automatically.
Run a single AWS CLI PutObjectTagging command with a bucket ARN as the object key and assume it tags every existing object under that bucket.
Use S3 Batch Operations with an appropriate object manifest, IAM permissions and completion report.

Explanation

Correct Answer

Batch Operations can apply supported actions at scale to enumerated object keys without per-object manual loops.

Incorrect Answers

Batch Operations acts on explicitly selected keys and versions from its manifest or supported inventory selection.

Storage Lens provides reporting; a supported Batch Operations job with explicit manifest and permissions is needed for changes.

PutObjectTagging targets particular objects; applying tags to millions of keys requires enumerating/manifesting those objects or an orchestrated supported bulk action.

Pass Your Amazon AWS Certified Solutions Architect - Associate SAA-C03 Exam Easy!

Question 650

Explanation

Correct Answer

Incorrect Answers

Question 651

Explanation

Correct Answer

Incorrect Answers

Question 652

Explanation

Correct Answer

Incorrect Answers

Question 653

Explanation

Correct Answer

Incorrect Answers

Question 654

Explanation

Correct Answer

Incorrect Answers

Question 655

Explanation

Correct Answer

Incorrect Answers

Question 656

Explanation

Correct Answer

Incorrect Answers

Question 657

Explanation

Correct Answer

Incorrect Answers

Question 658

Explanation

Correct Answer

Incorrect Answers

Question 659

Explanation

Correct Answer

Incorrect Answers

Question 660

Explanation

Correct Answer

Incorrect Answers

AWS Certified Solutions Architect - Associate SAA-C03 Practice Tests