Amazon AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 62

Question 610

Four independent analytics teams read the same Kinesis Data Stream and each needs a dedicated low-latency delivery lane without contending for shared shard-read bandwidth. Which feature?

Continue using shared GetRecords polling for each analytics team, assuming every reader automatically receives a separate dedicated 2 MiB/s per shard.
Increase the retention period from one to seven days while leaving all four applications on shared GetRecords polling.
Register separate Kinesis enhanced fan-out consumers for the independent applications.
Assign each analytics application a distinct IAM role but continue using the same shared shard polling bandwidth.

Explanation

Correct Answer

Each enhanced fan-out consumer gets dedicated per-shard throughput and push-based SubscribeToShard delivery.

Incorrect Answers

Enhanced fan-out consumers get their own throughput allocation; ordinary GetRecords consumers share the per-shard read throughput budget.

Longer retention widens replay time, but cannot allocate dedicated per-consumer read capacity when applications still share GetRecords shard throughput.

IAM identity isolation does not produce the dedicated per-consumer Kinesis read lane and HTTP/2 push behavior supplied by enhanced fan-out.

Question 611

Mobile events need continuous managed buffering, optional format transformation and delivery as objects into S3, without a custom long-running consumer application. What service fits?

Use an S3 Event Notification to transform each incoming device event before it reaches S3 without any ingestion service.
Use AWS Glue batch crawlers to classify any S3 files after arrival, without creating an incoming delivery stream.
Write the incoming records into an Amazon Kinesis Data Stream and assume the data stream itself automatically transforms each record and delivers consolidated S3 objects with no delivery consumer.
Use Amazon Data Firehose with an S3 delivery destination and configured buffering/transformations.

Explanation

Correct Answer

Firehose is managed event delivery into destinations such as S3; it differs from application-controlled Kinesis Streams retention.

Incorrect Answers

An S3 notification fires after an object is created, not as a managed record-buffering, optional transformation and delivery pipeline for raw events.

A Glue crawler can discover data already stored in S3 but is not the managed ingest/buffer/deliver pipeline for live mobile events.

Kinesis Data Streams captures and retains records for consumers. It does not itself supply the Firehose destination buffering and transformation pipeline; a consumer or an explicitly configured delivery integration is still needed.

Question 612

An S3 data lake receives CSV files each night; Athena users must query a new curated Parquet dataset, with the right partition columns. Select TWO actions for the pipeline.

Point the catalog table and partition metadata to the curated Parquet location and validate an Athena query over it.
Copy CSV objects under new S3 prefixes but keep their row formats, schema and partitions unchanged while calling them optimized Parquet.
Run a Glue crawler only against the original CSV files and expect the crawler to convert them into partitioned Parquet objects without a transformation job.
Transform the CSV inputs to correctly structured, partitioned Parquet using a supported AWS Glue ETL job.
Retain the catalog table pointing to the raw CSV prefix and assume Athena automatically discovers the separate Parquet output.

Explanation

Correct Answer

Creating new files is not enough if the consumers still query the old CSV table or stale partitions.

Glue can read source files, apply schema conversions and write columnar outputs that reduce Athena scanned bytes.

Incorrect Answers

Moving object names does not perform a file-format transformation or update Glue Data Catalog partitions for query pruning.

A crawler inventories and catalogs existing objects; it does not produce a newly transformed, correctly partitioned Parquet dataset. The requested ETL transformation and metadata update remain separate actions.

A query bound to the old catalog location continues scanning the raw source unless its metadata is changed.

Question 613

An analytics team needs to continuously transform arriving events from Kinesis into a curated S3 dataset, with checkpointed streaming state. Which AWS ETL approach?

Run one ordinary batch Glue job each morning without retaining offsets and assume it provides continuous streaming exactly-once sink semantics.
Run a supported AWS Glue streaming ETL job with appropriate checkpoints and incremental processing logic.
Enable Glue batch job bookmarks on a scheduled nightly job and assume they provide streaming checkpointed windows while the job is stopped.
Route the stream directly to a Glue crawler and assume catalog updates execute arbitrary Kinesis transformations continuously.

Explanation

Correct Answer

Glue streaming can consume ongoing events and maintain structured streaming checkpoints rather than rerunning full batches.

Incorrect Answers

Periodic batch ETL is not a substitute for continuous streaming with durable checkpoints and replay-safe sinks.

Batch job bookmarks track processed batch data; AWS Glue streaming ETL uses stream checkpoints and continuously running micro-batches rather than batch bookmark semantics.

Crawlers discover schema and metadata; they do not run continuous Spark streaming transforms or retain source-processing checkpoints.

Question 614

Analysts query partitioned files already stored in S3 once a week and want to run SQL without managing database servers. Which service?

Use Athena against unpartitioned CSV files and expect a WHERE predicate alone to prevent scanning unrelated source objects.
Query the registered S3 dataset with Amazon Athena and optimize partitions and columnar formats.
Keep the data in S3 but require a provisioned Redshift cluster to run every occasional SELECT against the source objects.
Run a Glue ETL job solely to execute each analyst SELECT and discard its outputs without using a query engine.

Explanation

Correct Answer

Athena performs serverless SQL over data in S3 using metadata, charging according to relevant query pricing.

Incorrect Answers

Athena is appropriate for serverless SQL, but partition pruning requires usable partition metadata and often efficient columnar format.

Redshift/Spectrum may be appropriate for other warehouse workloads; a serverless Athena query fits the stated intermittent SQL-on-S3 requirement without dedicated compute.

AWS Glue ETL prepares and transforms data; it is not a replacement for interactive serverless SQL over registered S3 tables.

Question 615

A BI warehouse repeatedly executes complex joins across billions of sales facts and dimensions with predictable daily concurrency. Which AWS engine supports those analytical scans?

Run each recurring full warehouse join directly on uncompressed row-formatted S3 CSV through ad hoc scans regardless of the measured workload.
Use an Athena-only ad hoc query pattern without benchmarking repeated joins, concurrency or a maintained warehouse schema.
Run one Glue crawler before each BI join and assume metadata scans provide a materialized join engine.
Use Amazon Redshift with a tested columnar schema and workload management design.

Explanation

Correct Answer

Redshift is purpose-built for recurring analytical scan/aggregate workloads; Athena suits a different serverless file-query pattern.

Incorrect Answers

S3/SQL-on-object query may fit occasional analytics, but repeated massive joins can be better served by a tuned warehouse with suitable data layout.

Athena can query lake files, but cannot be assumed optimal for predictable high-concurrency recurring warehouse joins without examining Redshift costs and workload management.

Crawlers maintain table/schema metadata but do not execute or optimize a recurring dimensional warehouse join workload.

Question 616

Three teams share one S3 data lake, but finance columns and tables must be accessible only to approved identities through governed analytical engines. Which layer provides centralized fine-grained governance?

Grant Lake Formation column restrictions but leave unrestricted S3 GetObject access to the raw underlying data for all consumers.
Rely only on the S3 bucket name prefix for finance files while granting all analytics principals direct raw-object reads.
Create one IAM role shared by every analyst with broad S3 GetObject and use application UI hiding to implement column security.
Use AWS Lake Formation permissions and catalog integration to govern supported lake tables/columns.

Explanation

Correct Answer

Lake Formation grants control access to registered catalog resources with supported services, complementing S3/IAM policies.

Incorrect Answers

Direct S3 permissions can bypass query-engine column restrictions; underlying storage IAM/bucket paths must also be restricted.

A naming convention without appropriately restricted S3 IAM policies and governed analytical permissions is not column/table-level access control.

A shared object-read principal may bypass analytical column restrictions and is not centralized fine-grained lake governance for separate teams.

Question 617

A file server holds 35 TiB of changing media files. The WAN stays connected, transfers must be encrypted, and the team needs incremental copy and checksum verification to EFS. Which managed service?

Schedule a full application-level copy over the WAN each night with no file verification and claim it is automatically incremental and consistent.
Use AWS DataSync with the supported source/destination locations and verification options.
Create an AWS Transfer Family SFTP endpoint and assume it automatically performs scheduled incremental NFS-to-EFS file comparisons and managed integrity verification.
Schedule repeated full-file rsync runs and declare the data validated solely because the copy process returned exit code zero.

Explanation

Correct Answer

DataSync provides scheduled/incremental data movement and integrity verification over connected networks.

Incorrect Answers

A separate copy script without comparison/verification does not deliver DataSync’s managed incremental transfer and integrity reporting.

AWS Transfer Family exposes supported file transfer protocols, but does not itself replace DataSync's source/destination file system comparison, incremental synchronization and configured integrity verification.

A custom workflow may transfer bytes but does not automatically provide the specified AWS managed incremental task and configured integrity verification checks.

Question 618

A new AWS customer with Enterprise Support has 700 TiB of portable storage at a remote research station, only intermittent 10-Mbps Internet, and a team that can transport the drives to an available AWS Data Transfer Terminal facility. The data must reach Amazon S3 in weeks. Which supported transfer plan best avoids months on the station WAN?

Reserve a supported AWS Data Transfer Terminal session, securely transport the customer-owned storage devices to that facility, and upload over its high-bandwidth connection to the approved AWS destination.
Transfer the 700-TiB archive only across the intermittent 10-Mbps link and assume multipart parallelism increases the physical line capacity.
Start AWS DataSync over the existing intermittent 10-Mbps research-station WAN and assume it automatically multiplies the physical connection bandwidth by a thousand.
Order a new AWS Snowball Edge device as a first-time Snow Family customer in October 2026 and assume the offline device will be shipped.

Explanation

Correct Answer

AWS Data Transfer Terminal provides a physical high-throughput upload location for supported Enterprise Support customers who bring their own drives. Confirm reservations, facility logistics and supported destination; it does not ship AWS Snow Family devices to new customers.

Incorrect Answers

More multipart workers may utilize available bandwidth but cannot increase the physical 10-Mbps uplink. A supported physical transfer facility or qualified partner path is needed for this newly onboarded customer's offline-sized deadline; new Snow Family device orders are not available.

DataSync can automate online transfers but cannot exceed an underlying 10-Mbps connection’s physical capacity for hundreds of TiB within the stated deadline.

AWS stopped offering new Snow Family device orders to new customers in November 2025; a new October 2026 customer should consider Data Transfer Terminal or eligible partners instead.

Question 619

An enterprise has existing Kafka clients that use Kafka broker APIs and needs a managed Kafka-compatible streaming cluster without rewriting application protocols. Which AWS service?

Replace Kafka clients with Kinesis producers without modifying broker protocols, assuming Kinesis exposes identical Kafka brokers.
Deploy Amazon Managed Streaming for Apache Kafka (Amazon MSK) with compatible brokers and security configuration.
Use a Kinesis Data Streams shard endpoint as the Kafka broker bootstrap address without a Kafka-compatible proxy or client changes.
Deploy an S3 event notification queue and treat it as a persistent Kafka topic broker with the same consumer group wire protocol.

Explanation

Correct Answer

MSK runs managed Kafka infrastructure and retains standard Kafka protocol compatibility for existing clients.

Incorrect Answers

Kinesis Data Streams is a different streaming API; existing Kafka-protocol clients need MSK or an explicitly designed compatibility migration.

Kinesis Data Streams exposes its own API, not the Kafka broker protocol expected by the legacy clients; Amazon MSK provides managed Kafka brokers.

Event notifications can signal object events, but the queue is not an Apache Kafka broker implementing client bootstrap/consumer group APIs.

Question 620

A production PostgreSQL database must remain online while rows are replicated continuously into a new AWS-managed database after an initial load. Which migration capability supports this?

Use AWS Database Migration Service full load plus change data capture, with supported engine and replication settings.
Run DMS in full-load-only mode during a live migration and assume committed source writes after the snapshot are automatically replayed.
Take a one-time database snapshot after the initial migration and assume it stays in sync with continuing PostgreSQL writes.
Enable the target database Multi-AZ failover and treat it as change-data capture from the separate source database.

Explanation

Correct Answer

DMS CDC processes ongoing source changes after the baseline copy, reducing migration downtime.

Incorrect Answers

Full load alone does not replicate ongoing changes; DMS requires compatible CDC configuration to keep the destination synchronized.

A static snapshot does not deliver new commits. DMS with supported ongoing CDC replication can apply changes after initial load.

Multi-AZ protects availability of a target deployment; it does not replicate transactions from the independent live source without a migration/CDC process.

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

Question 610

Explanation

Correct Answer

Incorrect Answers

Question 611

Explanation

Correct Answer

Incorrect Answers

Question 612

Explanation

Correct Answer

Incorrect Answers

Question 613

Explanation

Correct Answer

Incorrect Answers

Question 614

Explanation

Correct Answer

Incorrect Answers

Question 615

Explanation

Correct Answer

Incorrect Answers

Question 616

Explanation

Correct Answer

Incorrect Answers

Question 617

Explanation

Correct Answer

Incorrect Answers

Question 618

Explanation

Correct Answer

Incorrect Answers

Question 619

Explanation

Correct Answer

Incorrect Answers

Question 620

Explanation

Correct Answer

Incorrect Answers

AWS Certified Solutions Architect - Associate SAA-C03 Practice Tests