Amazon AWS Certified Solutions Architect Associate SAA-C03 Ingestion Streaming Transformation and Analytics Practice Test

 

Topic 15 covers ingestion streaming transformation and analytics for the AWS Certified Solutions Architect – Associate certification. These original practice questions apply the verified SAA-C03 objectives to practical decisions and troubleshooting. Select one answer unless a fixed number is requested. For broader preparation, visit the AWS Certified Solutions Architect Associate SAA-C03 Exam Dumps page. Each option includes an explanation of the relevant behavior and scenario constraints.

Question 1

A partner deposits one large CSV file every night and analytics can wait until the file arrives. Which ingestion style fits?

1. Run a scheduled ingestion task aligned to the hourly source availability.
2. Use AWS DMS with change data capture.
3. Use Amazon Kinesis Data Streams for the continuous event stream.
4. Use AWS Transfer Family with an S3 backend.
5. Use a scheduled batch ingestion workflow.

Correct Answer: E

 

Correct Answer

Answer E is correct because A complete file that arrives periodically does not require the complexity of a continuously running stream. This directly meets the decisive requirement: batch ingestion.

Incorrect Answers

Answer A is incorrect because Scheduled ingestion is appropriate when the upstream system only produces data at known intervals. It may be useful elsewhere, but it does not meet the decisive requirement here: batch ingestion.

Answer B is incorrect because DMS can perform an initial load and continue replicating source database changes until migration cutover. It may be useful elsewhere, but it does not meet the decisive requirement here: batch ingestion.

Answer C is incorrect because Kinesis Data Streams is designed for continuous high-throughput event ingestion with multiple consumers. It may be useful elsewhere, but it does not meet the decisive requirement here: batch ingestion.

Answer D is incorrect because Transfer Family provides managed SFTP and related protocol endpoints without operating customer-managed transfer servers. It may be useful elsewhere, but it does not meet the decisive requirement here: batch ingestion.

 

Question 2

An application emits thousands of click events per second and consumers need them within seconds. Which ingestion style and service fit?

1. Use a scheduled batch ingestion workflow.
2. Use AWS Transfer Family with an S3 backend.
3. Use Amazon Kinesis Data Streams for the continuous event stream.
4. Use AWS DMS with change data capture.
5. Run a scheduled ingestion task aligned to the hourly source availability.

Correct Answer: C

 

Correct Answer

Answer C is correct because Kinesis Data Streams is designed for continuous high-throughput event ingestion with multiple consumers. This directly meets the decisive requirement: stream with Kinesis.

Incorrect Answers

Answer A is incorrect because A complete file that arrives periodically does not require the complexity of a continuously running stream. It may be useful elsewhere, but it does not meet the decisive requirement here: stream with Kinesis.

Answer B is incorrect because Transfer Family provides managed SFTP and related protocol endpoints without operating customer-managed transfer servers. It may be useful elsewhere, but it does not meet the decisive requirement here: stream with Kinesis.

Answer D is incorrect because DMS can perform an initial load and continue replicating source database changes until migration cutover. It may be useful elsewhere, but it does not meet the decisive requirement here: stream with Kinesis.

Answer E is incorrect because Scheduled ingestion is appropriate when the upstream system only produces data at known intervals. It may be useful elsewhere, but it does not meet the decisive requirement here: stream with Kinesis.

 

Question 3

A database migration must capture ongoing inserts and updates after the initial load until cutover. Which AWS service capability fits?

1. Use a scheduled batch ingestion workflow.
2. Use AWS DMS with change data capture.
3. Use AWS Transfer Family with an S3 backend.
4. Run a scheduled ingestion task aligned to the hourly source availability.
5. Use Amazon Kinesis Data Streams for the continuous event stream.

Correct Answer: B

 

Correct Answer

Answer B is correct because DMS can perform an initial load and continue replicating source database changes until migration cutover. This directly meets the decisive requirement: DMS change data capture.

Incorrect Answers

Answer A is incorrect because A complete file that arrives periodically does not require the complexity of a continuously running stream. It may be useful elsewhere, but it does not meet the decisive requirement here: DMS change data capture.

Answer C is incorrect because Transfer Family provides managed SFTP and related protocol endpoints without operating customer-managed transfer servers. It may be useful elsewhere, but it does not meet the decisive requirement here: DMS change data capture.

Answer D is incorrect because Scheduled ingestion is appropriate when the upstream system only produces data at known intervals. It may be useful elsewhere, but it does not meet the decisive requirement here: DMS change data capture.

Answer E is incorrect because Kinesis Data Streams is designed for continuous high-throughput event ingestion with multiple consumers. It may be useful elsewhere, but it does not meet the decisive requirement here: DMS change data capture.

 

Question 4

A SaaS export is available once per hour through a scheduled API call. Which design matches the source cadence?

1. Use Amazon Kinesis Data Streams for the continuous event stream.
2. Use AWS Transfer Family with an S3 backend.
3. Use AWS DMS with change data capture.
4. Use a scheduled batch ingestion workflow.
5. Run a scheduled ingestion task aligned to the hourly source availability.

Correct Answer: E

 

Correct Answer

Answer E is correct because Scheduled ingestion is appropriate when the upstream system only produces data at known intervals. This directly meets the decisive requirement: scheduled ingestion job.

Incorrect Answers

Answer A is incorrect because Kinesis Data Streams is designed for continuous high-throughput event ingestion with multiple consumers. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled ingestion job.

Answer B is incorrect because Transfer Family provides managed SFTP and related protocol endpoints without operating customer-managed transfer servers. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled ingestion job.

Answer C is incorrect because DMS can perform an initial load and continue replicating source database changes until migration cutover. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled ingestion job.

Answer D is incorrect because A complete file that arrives periodically does not require the complexity of a continuously running stream. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled ingestion job.

 

Question 5

External partners require managed SFTP uploads into S3. Which AWS service should provide the ingress endpoint?

1. Use AWS Transfer Family with an S3 backend.
2. Use AWS DMS with change data capture.
3. Use a scheduled batch ingestion workflow.
4. Run a scheduled ingestion task aligned to the hourly source availability.
5. Use Amazon Kinesis Data Streams for the continuous event stream.

Correct Answer: A

 

Correct Answer

Answer A is correct because Transfer Family provides managed SFTP and related protocol endpoints without operating customer-managed transfer servers. This directly meets the decisive requirement: AWS Transfer Family.

Incorrect Answers

Answer B is incorrect because DMS can perform an initial load and continue replicating source database changes until migration cutover. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Transfer Family.

Answer C is incorrect because A complete file that arrives periodically does not require the complexity of a continuously running stream. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Transfer Family.

Answer D is incorrect because Scheduled ingestion is appropriate when the upstream system only produces data at known intervals. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Transfer Family.

Answer E is incorrect because Kinesis Data Streams is designed for continuous high-throughput event ingestion with multiple consumers. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Transfer Family.

 

Question 6

Events for one account must be processed in order, while other accounts can be processed in parallel. Which Kinesis design element should group related records?

1. Scale or optimize consumer processing.
2. Use the stream retention window and consumer checkpoints to replay retained records.
3. Choose a partition key that groups records needing order while distributing unrelated entities across shards.
4. Use independent consumers for each application.
5. Change the partition-key strategy and reshard if required.

Correct Answer: C

 

Correct Answer

Answer C is correct because Kinesis preserves record order within a shard, so partition-key design controls both ordering and load distribution. This directly meets the decisive requirement: partition key.

Incorrect Answers

Answer A is incorrect because Growing iterator age indicates consumers are falling behind the incoming stream rather than the producer being unable to write records. It may be useful elsewhere, but it does not meet the decisive requirement here: partition key.

Answer B is incorrect because Kinesis retains records for the configured period so consumers can reread data that has not expired. It may be useful elsewhere, but it does not meet the decisive requirement here: partition key.

Answer D is incorrect because Multiple consumers can read the same stream independently without using a work-queue model where one consumer removes the record for others. It may be useful elsewhere, but it does not meet the decisive requirement here: partition key.

Answer E is incorrect because A hot key concentrates traffic on one shard even when aggregate capacity exists on other shards. It may be useful elsewhere, but it does not meet the decisive requirement here: partition key.

 

Question 7

Two independent applications must process the same Kinesis stream without consuming each other’s copy of the data. Which design is appropriate?

1. Choose a partition key that groups records needing order while distributing unrelated entities across shards.
2. Change the partition-key strategy and reshard if required.
3. Scale or optimize consumer processing.
4. Use the stream retention window and consumer checkpoints to replay retained records.
5. Use independent consumers for each application.

Correct Answer: E

 

Correct Answer

Answer E is correct because Multiple consumers can read the same stream independently without using a work-queue model where one consumer removes the record for others. This directly meets the decisive requirement: separate stream consumers.

Incorrect Answers

Answer A is incorrect because Kinesis preserves record order within a shard, so partition-key design controls both ordering and load distribution. It may be useful elsewhere, but it does not meet the decisive requirement here: separate stream consumers.

Answer B is incorrect because A hot key concentrates traffic on one shard even when aggregate capacity exists on other shards. It may be useful elsewhere, but it does not meet the decisive requirement here: separate stream consumers.

Answer C is incorrect because Growing iterator age indicates consumers are falling behind the incoming stream rather than the producer being unable to write records. It may be useful elsewhere, but it does not meet the decisive requirement here: separate stream consumers.

Answer D is incorrect because Kinesis retains records for the configured period so consumers can reread data that has not expired. It may be useful elsewhere, but it does not meet the decisive requirement here: separate stream consumers.

 

Question 8

A bug in a consumer is fixed and the team needs to reprocess records that are still within the stream retention window. Which property enables this?

1. Choose a partition key that groups records needing order while distributing unrelated entities across shards.
2. Use independent consumers for each application.
3. Scale or optimize consumer processing.
4. Change the partition-key strategy and reshard if required.
5. Use the stream retention window and consumer checkpoints to replay retained records.

Correct Answer: E

 

Correct Answer

Answer E is correct because Kinesis retains records for the configured period so consumers can reread data that has not expired. This directly meets the decisive requirement: re-read records by sequence/checkpoint.

Incorrect Answers

Answer A is incorrect because Kinesis preserves record order within a shard, so partition-key design controls both ordering and load distribution. It may be useful elsewhere, but it does not meet the decisive requirement here: re-read records by sequence/checkpoint.

Answer B is incorrect because Multiple consumers can read the same stream independently without using a work-queue model where one consumer removes the record for others. It may be useful elsewhere, but it does not meet the decisive requirement here: re-read records by sequence/checkpoint.

Answer C is incorrect because Growing iterator age indicates consumers are falling behind the incoming stream rather than the producer being unable to write records. It may be useful elsewhere, but it does not meet the decisive requirement here: re-read records by sequence/checkpoint.

Answer D is incorrect because A hot key concentrates traffic on one shard even when aggregate capacity exists on other shards. It may be useful elsewhere, but it does not meet the decisive requirement here: re-read records by sequence/checkpoint.

 

Question 9

One Kinesis shard is overloaded while others are idle because most events use the same partition key. What should change?

1. Change the partition-key strategy and reshard if required.
2. Use the stream retention window and consumer checkpoints to replay retained records.
3. Scale or optimize consumer processing.
4. Choose a partition key that groups records needing order while distributing unrelated entities across shards.
5. Use independent consumers for each application.

Correct Answer: A

 

Correct Answer

Answer A is correct because A hot key concentrates traffic on one shard even when aggregate capacity exists on other shards. This directly meets the decisive requirement: redistribute partition keys/reshard.

Incorrect Answers

Answer B is incorrect because Kinesis retains records for the configured period so consumers can reread data that has not expired. It may be useful elsewhere, but it does not meet the decisive requirement here: redistribute partition keys/reshard.

Answer C is incorrect because Growing iterator age indicates consumers are falling behind the incoming stream rather than the producer being unable to write records. It may be useful elsewhere, but it does not meet the decisive requirement here: redistribute partition keys/reshard.

Answer D is incorrect because Kinesis preserves record order within a shard, so partition-key design controls both ordering and load distribution. It may be useful elsewhere, but it does not meet the decisive requirement here: redistribute partition keys/reshard.

Answer E is incorrect because Multiple consumers can read the same stream independently without using a work-queue model where one consumer removes the record for others. It may be useful elsewhere, but it does not meet the decisive requirement here: redistribute partition keys/reshard.

 

Question 10

Incoming records are current, but consumer iterator age keeps rising. Which resource is most likely insufficient?

1. Change the partition-key strategy and reshard if required.
2. Choose a partition key that groups records needing order while distributing unrelated entities across shards.
3. Use independent consumers for each application.
4. Scale or optimize consumer processing.
5. Use the stream retention window and consumer checkpoints to replay retained records.

Correct Answer: D

 

Correct Answer

Answer D is correct because Growing iterator age indicates consumers are falling behind the incoming stream rather than the producer being unable to write records. This directly meets the decisive requirement: consumer processing capacity.

Incorrect Answers

Answer A is incorrect because A hot key concentrates traffic on one shard even when aggregate capacity exists on other shards. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer B is incorrect because Kinesis preserves record order within a shard, so partition-key design controls both ordering and load distribution. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer C is incorrect because Multiple consumers can read the same stream independently without using a work-queue model where one consumer removes the record for others. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer E is incorrect because Kinesis retains records for the configured period so consumers can reread data that has not expired. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

 

Question 11

A company synchronizes 30 TB of changed files from on premises to S3 every weekend over a high-bandwidth link. Which managed service is appropriate?

1. Verify the current AWS offline-transfer service availability and account eligibility before designing around physical devices.
2. Use AWS Database Migration Service.
3. Use AWS DataSync for the recurring online file transfer.
4. Use AWS Storage Gateway File Gateway.
5. Schedule incremental DataSync jobs during the allowed bandwidth window.

Correct Answer: C

 

Correct Answer

Answer C is correct because DataSync manages high-throughput file/object transfer and supports repeated synchronization jobs over available network connectivity. This directly meets the decisive requirement: DataSync.

Incorrect Answers

Answer A is incorrect because Snowball is no longer available to new customers, so current physical-transfer options and lifecycle constraints must be checked instead of assuming a Snow device can be ordered. It may be useful elsewhere, but it does not meet the decisive requirement here: DataSync.

Answer B is incorrect because DMS is designed for database migration and can replicate ongoing source changes to reduce cutover downtime. It may be useful elsewhere, but it does not meet the decisive requirement here: DataSync.

Answer D is incorrect because File Gateway provides local file-protocol access while storing durable objects in S3 and caching active data on premises. It may be useful elsewhere, but it does not meet the decisive requirement here: DataSync.

Answer E is incorrect because Incremental managed transfer can use the available window repeatedly without retransferring unchanged files. It may be useful elsewhere, but it does not meet the decisive requirement here: DataSync.

 

Question 12

On-premises applications need local file access while objects are durably stored in S3 with a cache at the site. Which family of service fits?

1. Use AWS Database Migration Service.
2. Verify the current AWS offline-transfer service availability and account eligibility before designing around physical devices.
3. Schedule incremental DataSync jobs during the allowed bandwidth window.
4. Use AWS DataSync for the recurring online file transfer.
5. Use AWS Storage Gateway File Gateway.

Correct Answer: E

 

Correct Answer

Answer E is correct because File Gateway provides local file-protocol access while storing durable objects in S3 and caching active data on premises. This directly meets the decisive requirement: Storage Gateway File Gateway.

Incorrect Answers

Answer A is incorrect because DMS is designed for database migration and can replicate ongoing source changes to reduce cutover downtime. It may be useful elsewhere, but it does not meet the decisive requirement here: Storage Gateway File Gateway.

Answer B is incorrect because Snowball is no longer available to new customers, so current physical-transfer options and lifecycle constraints must be checked instead of assuming a Snow device can be ordered. It may be useful elsewhere, but it does not meet the decisive requirement here: Storage Gateway File Gateway.

Answer C is incorrect because Incremental managed transfer can use the available window repeatedly without retransferring unchanged files. It may be useful elsewhere, but it does not meet the decisive requirement here: Storage Gateway File Gateway.

Answer D is incorrect because DataSync manages high-throughput file/object transfer and supports repeated synchronization jobs over available network connectivity. It may be useful elsewhere, but it does not meet the decisive requirement here: Storage Gateway File Gateway.

 

Question 13

A new customer has a multi-petabyte migration and the available WAN would take months. Which current planning step is essential before choosing an offline device option?

1. Use AWS Database Migration Service.
2. Schedule incremental DataSync jobs during the allowed bandwidth window.
3. Use AWS DataSync for the recurring online file transfer.
4. Use AWS Storage Gateway File Gateway.
5. Verify the current AWS offline-transfer service availability and account eligibility before designing around physical devices.

Correct Answer: E

 

Correct Answer

Answer E is correct because Snowball is no longer available to new customers, so current physical-transfer options and lifecycle constraints must be checked instead of assuming a Snow device can be ordered. This directly meets the decisive requirement: verify current AWS physical-transfer options because Snowball is unavailable to new customers.

Incorrect Answers

Answer A is incorrect because DMS is designed for database migration and can replicate ongoing source changes to reduce cutover downtime. It may be useful elsewhere, but it does not meet the decisive requirement here: verify current AWS physical-transfer options because Snowball is unavailable to new customers.

Answer B is incorrect because Incremental managed transfer can use the available window repeatedly without retransferring unchanged files. It may be useful elsewhere, but it does not meet the decisive requirement here: verify current AWS physical-transfer options because Snowball is unavailable to new customers.

Answer C is incorrect because DataSync manages high-throughput file/object transfer and supports repeated synchronization jobs over available network connectivity. It may be useful elsewhere, but it does not meet the decisive requirement here: verify current AWS physical-transfer options because Snowball is unavailable to new customers.

Answer D is incorrect because File Gateway provides local file-protocol access while storing durable objects in S3 and caching active data on premises. It may be useful elsewhere, but it does not meet the decisive requirement here: verify current AWS physical-transfer options because Snowball is unavailable to new customers.

 

Question 14

A relational database must move to AWS with minimal downtime and ongoing changes replicated until cutover. Which service is designed for this?

1. Use AWS Database Migration Service.
2. Use AWS Storage Gateway File Gateway.
3. Schedule incremental DataSync jobs during the allowed bandwidth window.
4. Use AWS DataSync for the recurring online file transfer.
5. Verify the current AWS offline-transfer service availability and account eligibility before designing around physical devices.

Correct Answer: A

 

Correct Answer

Answer A is correct because DMS is designed for database migration and can replicate ongoing source changes to reduce cutover downtime. This directly meets the decisive requirement: AWS DMS.

Incorrect Answers

Answer B is incorrect because File Gateway provides local file-protocol access while storing durable objects in S3 and caching active data on premises. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS DMS.

Answer C is incorrect because Incremental managed transfer can use the available window repeatedly without retransferring unchanged files. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS DMS.

Answer D is incorrect because DataSync manages high-throughput file/object transfer and supports repeated synchronization jobs over available network connectivity. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS DMS.

Answer E is incorrect because Snowball is no longer available to new customers, so current physical-transfer options and lifecycle constraints must be checked instead of assuming a Snow device can be ordered. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS DMS.

 

Question 15

A site has only a four-hour high-bandwidth network window each night. Which transfer design best uses the window?

1. Schedule incremental DataSync jobs during the allowed bandwidth window.
2. Verify the current AWS offline-transfer service availability and account eligibility before designing around physical devices.
3. Use AWS Database Migration Service.
4. Use AWS Storage Gateway File Gateway.
5. Use AWS DataSync for the recurring online file transfer.

Correct Answer: A

 

Correct Answer

Answer A is correct because Incremental managed transfer can use the available window repeatedly without retransferring unchanged files. This directly meets the decisive requirement: scheduled incremental DataSync during the window.

Incorrect Answers

Answer B is incorrect because Snowball is no longer available to new customers, so current physical-transfer options and lifecycle constraints must be checked instead of assuming a Snow device can be ordered. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled incremental DataSync during the window.

Answer C is incorrect because DMS is designed for database migration and can replicate ongoing source changes to reduce cutover downtime. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled incremental DataSync during the window.

Answer D is incorrect because File Gateway provides local file-protocol access while storing durable objects in S3 and caching active data on premises. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled incremental DataSync during the window.

Answer E is incorrect because DataSync manages high-throughput file/object transfer and supports repeated synchronization jobs over available network connectivity. It may be useful elsewhere, but it does not meet the decisive requirement here: scheduled incremental DataSync during the window.

 

Question 16

A data lake receives CSV files and must convert them to partitioned Parquet on a schedule without managing servers. Which service is the best fit?

1. Use AWS Lambda for the short event-driven transformation.
2. Use Amazon Athena with SQL such as CTAS where the transformation fits serverless SQL.
3. Use AWS Glue and the Glue Data Catalog.
4. Use AWS Glue ETL.
5. Use Amazon EMR for the Spark workload.

Correct Answer: D

 

Correct Answer

Answer D is correct because Glue provides managed/serverless data integration and ETL capabilities and can convert data formats such as CSV to Parquet. This directly meets the decisive requirement: AWS Glue.

Incorrect Answers

Answer A is incorrect because Lambda is suited to short stateless transformations triggered by object or event arrival. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer B is incorrect because Athena can query S3 directly and write query results or transformed datasets without managing a database server. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer C is incorrect because Glue combines ETL with a managed metadata catalog for data-lake schemas and transformed datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer E is incorrect because EMR provides managed clusters for frameworks such as Spark when the workload needs cluster/runtime control beyond a simple ETL job. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

 

Question 17

A team already has a large Spark job with custom libraries and needs cluster-level control over the Spark runtime. Which service fits better than a simple serverless ETL job?

1. Use AWS Lambda for the short event-driven transformation.
2. Use AWS Glue ETL.
3. Use AWS Glue and the Glue Data Catalog.
4. Use Amazon Athena with SQL such as CTAS where the transformation fits serverless SQL.
5. Use Amazon EMR for the Spark workload.

Correct Answer: E

 

Correct Answer

Answer E is correct because EMR provides managed clusters for frameworks such as Spark when the workload needs cluster/runtime control beyond a simple ETL job. This directly meets the decisive requirement: Amazon EMR.

Incorrect Answers

Answer A is incorrect because Lambda is suited to short stateless transformations triggered by object or event arrival. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon EMR.

Answer B is incorrect because Glue provides managed/serverless data integration and ETL capabilities and can convert data formats such as CSV to Parquet. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon EMR.

Answer C is incorrect because Glue combines ETL with a managed metadata catalog for data-lake schemas and transformed datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon EMR.

Answer D is incorrect because Athena can query S3 directly and write query results or transformed datasets without managing a database server. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon EMR.

 

Question 18

Each uploaded object requires a small transformation that completes in seconds before metadata is written. Which compute pattern is appropriate?

1. Use Amazon EMR for the Spark workload.
2. Use AWS Glue and the Glue Data Catalog.
3. Use AWS Glue ETL.
4. Use Amazon Athena with SQL such as CTAS where the transformation fits serverless SQL.
5. Use AWS Lambda for the short event-driven transformation.

Correct Answer: E

 

Correct Answer

Answer E is correct because Lambda is suited to short stateless transformations triggered by object or event arrival. This directly meets the decisive requirement: Lambda event processing.

Incorrect Answers

Answer A is incorrect because EMR provides managed clusters for frameworks such as Spark when the workload needs cluster/runtime control beyond a simple ETL job. It may be useful elsewhere, but it does not meet the decisive requirement here: Lambda event processing.

Answer B is incorrect because Glue combines ETL with a managed metadata catalog for data-lake schemas and transformed datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: Lambda event processing.

Answer C is incorrect because Glue provides managed/serverless data integration and ETL capabilities and can convert data formats such as CSV to Parquet. It may be useful elsewhere, but it does not meet the decisive requirement here: Lambda event processing.

Answer D is incorrect because Athena can query S3 directly and write query results or transformed datasets without managing a database server. It may be useful elsewhere, but it does not meet the decisive requirement here: Lambda event processing.

 

Question 19

Analysts need to create a transformed table from S3 using SQL and the result can be generated as a query workflow rather than a long-running ETL cluster. Which service is suitable?

1. Use AWS Glue ETL.
2. Use AWS Glue and the Glue Data Catalog.
3. Use Amazon Athena with SQL such as CTAS where the transformation fits serverless SQL.
4. Use AWS Lambda for the short event-driven transformation.
5. Use Amazon EMR for the Spark workload.

Correct Answer: C

 

Correct Answer

Answer C is correct because Athena can query S3 directly and write query results or transformed datasets without managing a database server. This directly meets the decisive requirement: Athena CTAS/query.

Incorrect Answers

Answer A is incorrect because Glue provides managed/serverless data integration and ETL capabilities and can convert data formats such as CSV to Parquet. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena CTAS/query.

Answer B is incorrect because Glue combines ETL with a managed metadata catalog for data-lake schemas and transformed datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena CTAS/query.

Answer D is incorrect because Lambda is suited to short stateless transformations triggered by object or event arrival. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena CTAS/query.

Answer E is incorrect because EMR provides managed clusters for frameworks such as Spark when the workload needs cluster/runtime control beyond a simple ETL job. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena CTAS/query.

 

Question 20

A pipeline repeatedly converts semi-structured records to columnar format and maintains a centralized schema catalog. Which managed service is a natural fit?

1. Use AWS Glue ETL.
2. Use Amazon Athena with SQL such as CTAS where the transformation fits serverless SQL.
3. Use AWS Glue and the Glue Data Catalog.
4. Use AWS Lambda for the short event-driven transformation.
5. Use Amazon EMR for the Spark workload.

Correct Answer: C

 

Correct Answer

Answer C is correct because Glue combines ETL with a managed metadata catalog for data-lake schemas and transformed datasets. This directly meets the decisive requirement: AWS Glue.

Incorrect Answers

Answer A is incorrect because Glue provides managed/serverless data integration and ETL capabilities and can convert data formats such as CSV to Parquet. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer B is incorrect because Athena can query S3 directly and write query results or transformed datasets without managing a database server. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer D is incorrect because Lambda is suited to short stateless transformations triggered by object or event arrival. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

Answer E is incorrect because EMR provides managed clusters for frameworks such as Spark when the workload needs cluster/runtime control beyond a simple ETL job. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Glue.

 

Question 21

A central S3 data lake contains finance and marketing datasets, and analysts should see only authorized columns/tables. Which AWS governance service can centrally manage fine-grained data-lake permissions?

1. Keep metadata/catalog permissions separate from permission to read the underlying S3 objects.
2. Use governed cross-account sharing or delegated read access to the selected datasets.
3. Use a tightly scoped cross-account role or bucket policy that permits only the required object writes.
4. Use AWS Lake Formation for governed data-lake permissions where its supported resources fit the design.
5. Encrypt S3 data with AWS KMS keys whose policies and grants restrict cryptographic use to authorized principals.

Correct Answer: D

 

Correct Answer

Answer D is correct because Lake Formation can centralize fine-grained permissions over registered data-lake resources and metadata. This directly meets the decisive requirement: Lake Formation.

Incorrect Answers

Answer A is incorrect because Knowing a table schema does not require broad object-data access, and the two authorization layers should be scoped independently. It may be useful elsewhere, but it does not meet the decisive requirement here: Lake Formation.

Answer B is incorrect because Centralized access avoids uncontrolled copies while allowing consumers to read only the data products they are authorized to use. It may be useful elsewhere, but it does not meet the decisive requirement here: Lake Formation.

Answer C is incorrect because A producer can be granted the minimum write permissions without gaining read or administrative access to unrelated lake data. It may be useful elsewhere, but it does not meet the decisive requirement here: Lake Formation.

Answer E is incorrect because SSE-KMS provides encryption at rest with customer-controlled key access policy boundaries. It may be useful elsewhere, but it does not meet the decisive requirement here: Lake Formation.

 

Question 22

A Glue crawler needs to discover schema but should not grant analysts direct access to the underlying S3 objects. Which security principle applies?

1. Keep metadata/catalog permissions separate from permission to read the underlying S3 objects.
2. Use a tightly scoped cross-account role or bucket policy that permits only the required object writes.
3. Use governed cross-account sharing or delegated read access to the selected datasets.
4. Use AWS Lake Formation for governed data-lake permissions where its supported resources fit the design.
5. Encrypt S3 data with AWS KMS keys whose policies and grants restrict cryptographic use to authorized principals.

Correct Answer: A

 

Correct Answer

Answer A is correct because Knowing a table schema does not require broad object-data access, and the two authorization layers should be scoped independently. This directly meets the decisive requirement: separate catalog permissions from S3 data permissions.

Incorrect Answers

Answer B is incorrect because A producer can be granted the minimum write permissions without gaining read or administrative access to unrelated lake data. It may be useful elsewhere, but it does not meet the decisive requirement here: separate catalog permissions from S3 data permissions.

Answer C is incorrect because Centralized access avoids uncontrolled copies while allowing consumers to read only the data products they are authorized to use. It may be useful elsewhere, but it does not meet the decisive requirement here: separate catalog permissions from S3 data permissions.

Answer D is incorrect because Lake Formation can centralize fine-grained permissions over registered data-lake resources and metadata. It may be useful elsewhere, but it does not meet the decisive requirement here: separate catalog permissions from S3 data permissions.

Answer E is incorrect because SSE-KMS provides encryption at rest with customer-controlled key access policy boundaries. It may be useful elsewhere, but it does not meet the decisive requirement here: separate catalog permissions from S3 data permissions.

 

Question 23

Objects in the lake contain regulated data and must be encrypted at rest with customer-controlled access. Which control should be used?

1. Use a tightly scoped cross-account role or bucket policy that permits only the required object writes.
2. Use governed cross-account sharing or delegated read access to the selected datasets.
3. Encrypt S3 data with AWS KMS keys whose policies and grants restrict cryptographic use to authorized principals.
4. Keep metadata/catalog permissions separate from permission to read the underlying S3 objects.
5. Use AWS Lake Formation for governed data-lake permissions where its supported resources fit the design.

Correct Answer: C

 

Correct Answer

Answer C is correct because SSE-KMS provides encryption at rest with customer-controlled key access policy boundaries. This directly meets the decisive requirement: SSE-KMS with managed key policy.

Incorrect Answers

Answer A is incorrect because A producer can be granted the minimum write permissions without gaining read or administrative access to unrelated lake data. It may be useful elsewhere, but it does not meet the decisive requirement here: SSE-KMS with managed key policy.

Answer B is incorrect because Centralized access avoids uncontrolled copies while allowing consumers to read only the data products they are authorized to use. It may be useful elsewhere, but it does not meet the decisive requirement here: SSE-KMS with managed key policy.

Answer D is incorrect because Knowing a table schema does not require broad object-data access, and the two authorization layers should be scoped independently. It may be useful elsewhere, but it does not meet the decisive requirement here: SSE-KMS with managed key policy.

Answer E is incorrect because Lake Formation can centralize fine-grained permissions over registered data-lake resources and metadata. It may be useful elsewhere, but it does not meet the decisive requirement here: SSE-KMS with managed key policy.

 

Question 24

Data producers from another account must deliver objects to a central lake but must not gain read access to other datasets. Which design is appropriate?

1. Use AWS Lake Formation for governed data-lake permissions where its supported resources fit the design.
2. Encrypt S3 data with AWS KMS keys whose policies and grants restrict cryptographic use to authorized principals.
3. Keep metadata/catalog permissions separate from permission to read the underlying S3 objects.
4. Use governed cross-account sharing or delegated read access to the selected datasets.
5. Use a tightly scoped cross-account role or bucket policy that permits only the required object writes.

Correct Answer: E

 

Correct Answer

Answer E is correct because A producer can be granted the minimum write permissions without gaining read or administrative access to unrelated lake data. This directly meets the decisive requirement: scoped cross-account role/bucket policy.

Incorrect Answers

Answer A is incorrect because Lake Formation can centralize fine-grained permissions over registered data-lake resources and metadata. It may be useful elsewhere, but it does not meet the decisive requirement here: scoped cross-account role/bucket policy.

Answer B is incorrect because SSE-KMS provides encryption at rest with customer-controlled key access policy boundaries. It may be useful elsewhere, but it does not meet the decisive requirement here: scoped cross-account role/bucket policy.

Answer C is incorrect because Knowing a table schema does not require broad object-data access, and the two authorization layers should be scoped independently. It may be useful elsewhere, but it does not meet the decisive requirement here: scoped cross-account role/bucket policy.

Answer D is incorrect because Centralized access avoids uncontrolled copies while allowing consumers to read only the data products they are authorized to use. It may be useful elsewhere, but it does not meet the decisive requirement here: scoped cross-account role/bucket policy.

 

Question 25

A consumer account needs governed read access to selected data products without copying the entire lake. Which design best preserves centralized governance?

1. Keep metadata/catalog permissions separate from permission to read the underlying S3 objects.
2. Use governed cross-account sharing or delegated read access to the selected datasets.
3. Use AWS Lake Formation for governed data-lake permissions where its supported resources fit the design.
4. Encrypt S3 data with AWS KMS keys whose policies and grants restrict cryptographic use to authorized principals.
5. Use a tightly scoped cross-account role or bucket policy that permits only the required object writes.

Correct Answer: B

 

Correct Answer

Answer B is correct because Centralized access avoids uncontrolled copies while allowing consumers to read only the data products they are authorized to use. This directly meets the decisive requirement: cross-account governed sharing/access.

Incorrect Answers

Answer A is incorrect because Knowing a table schema does not require broad object-data access, and the two authorization layers should be scoped independently. It may be useful elsewhere, but it does not meet the decisive requirement here: cross-account governed sharing/access.

Answer C is incorrect because Lake Formation can centralize fine-grained permissions over registered data-lake resources and metadata. It may be useful elsewhere, but it does not meet the decisive requirement here: cross-account governed sharing/access.

Answer D is incorrect because SSE-KMS provides encryption at rest with customer-controlled key access policy boundaries. It may be useful elsewhere, but it does not meet the decisive requirement here: cross-account governed sharing/access.

Answer E is incorrect because A producer can be granted the minimum write permissions without gaining read or administrative access to unrelated lake data. It may be useful elsewhere, but it does not meet the decisive requirement here: cross-account governed sharing/access.

 

Question 26

Analysts run occasional SQL queries directly against Parquet files in S3 and do not want to manage a warehouse cluster. Which service is appropriate?

1. Use AWS Data Exchange for governed subscription and distribution of supported third-party data products.
2. Use Amazon Athena.
3. Use Amazon Quick for managed business intelligence and visualization where it meets the dashboard requirements.
4. Use Amazon OpenSearch Service.
5. Use Amazon Redshift.

Correct Answer: B

 

Correct Answer

Answer B is correct because Athena provides serverless SQL queries over data in S3 and is well suited to ad-hoc analytical access without managing database infrastructure. This directly meets the decisive requirement: Athena.

Incorrect Answers

Answer A is incorrect because Data Exchange is designed to discover, subscribe to, and use third-party data products rather than building a custom data marketplace from scratch. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena.

Answer C is incorrect because The current SAA-C03 scope lists Amazon Quick as an analytics and visualization service; it is intended for managed BI dashboards. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena.

Answer D is incorrect because OpenSearch is suited to indexed full-text search, log analytics, and dashboard-oriented exploration. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena.

Answer E is incorrect because Redshift is a managed analytical warehouse designed for complex SQL across large curated datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: Athena.

 

Question 27

A BI workload runs complex repeated joins over a large curated analytical model and needs warehouse performance. Which service is appropriate?

1. Use Amazon Quick for managed business intelligence and visualization where it meets the dashboard requirements.
2. Use Amazon Athena.
3. Use AWS Data Exchange for governed subscription and distribution of supported third-party data products.
4. Use Amazon Redshift.
5. Use Amazon OpenSearch Service.

Correct Answer: D

 

Correct Answer

Answer D is correct because Redshift is a managed analytical warehouse designed for complex SQL across large curated datasets. This directly meets the decisive requirement: Redshift.

Incorrect Answers

Answer A is incorrect because The current SAA-C03 scope lists Amazon Quick as an analytics and visualization service; it is intended for managed BI dashboards. It may be useful elsewhere, but it does not meet the decisive requirement here: Redshift.

Answer B is incorrect because Athena provides serverless SQL queries over data in S3 and is well suited to ad-hoc analytical access without managing database infrastructure. It may be useful elsewhere, but it does not meet the decisive requirement here: Redshift.

Answer C is incorrect because Data Exchange is designed to discover, subscribe to, and use third-party data products rather than building a custom data marketplace from scratch. It may be useful elsewhere, but it does not meet the decisive requirement here: Redshift.

Answer E is incorrect because OpenSearch is suited to indexed full-text search, log analytics, and dashboard-oriented exploration. It may be useful elsewhere, but it does not meet the decisive requirement here: Redshift.

 

Question 28

Operations teams need full-text search and near-real-time log exploration with dashboards. Which service is a good fit?

1. Use Amazon Quick for managed business intelligence and visualization where it meets the dashboard requirements.
2. Use Amazon Redshift.
3. Use Amazon Athena.
4. Use AWS Data Exchange for governed subscription and distribution of supported third-party data products.
5. Use Amazon OpenSearch Service.

Correct Answer: E

 

Correct Answer

Answer E is correct because OpenSearch is suited to indexed full-text search, log analytics, and dashboard-oriented exploration. This directly meets the decisive requirement: OpenSearch.

Incorrect Answers

Answer A is incorrect because The current SAA-C03 scope lists Amazon Quick as an analytics and visualization service; it is intended for managed BI dashboards. It may be useful elsewhere, but it does not meet the decisive requirement here: OpenSearch.

Answer B is incorrect because Redshift is a managed analytical warehouse designed for complex SQL across large curated datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: OpenSearch.

Answer C is incorrect because Athena provides serverless SQL queries over data in S3 and is well suited to ad-hoc analytical access without managing database infrastructure. It may be useful elsewhere, but it does not meet the decisive requirement here: OpenSearch.

Answer D is incorrect because Data Exchange is designed to discover, subscribe to, and use third-party data products rather than building a custom data marketplace from scratch. It may be useful elsewhere, but it does not meet the decisive requirement here: OpenSearch.

 

Question 29

Executives need managed interactive dashboards over curated datasets. Which AWS analytics/BI service should the architecture consider according to current exam scope?

1. Use AWS Data Exchange for governed subscription and distribution of supported third-party data products.
2. Use Amazon OpenSearch Service.
3. Use Amazon Redshift.
4. Use Amazon Quick for managed business intelligence and visualization where it meets the dashboard requirements.
5. Use Amazon Athena.

Correct Answer: D

 

Correct Answer

Answer D is correct because The current SAA-C03 scope lists Amazon Quick as an analytics and visualization service; it is intended for managed BI dashboards. This directly meets the decisive requirement: Amazon Quick.

Incorrect Answers

Answer A is incorrect because Data Exchange is designed to discover, subscribe to, and use third-party data products rather than building a custom data marketplace from scratch. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon Quick.

Answer B is incorrect because OpenSearch is suited to indexed full-text search, log analytics, and dashboard-oriented exploration. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon Quick.

Answer C is incorrect because Redshift is a managed analytical warehouse designed for complex SQL across large curated datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon Quick.

Answer E is incorrect because Athena provides serverless SQL queries over data in S3 and is well suited to ad-hoc analytical access without managing database infrastructure. It may be useful elsewhere, but it does not meet the decisive requirement here: Amazon Quick.

 

Question 30

A business wants to publish curated third-party datasets to internal consumers with governed subscription/access patterns. Which in-scope AWS service category fits?

1. Use Amazon Redshift.
2. Use Amazon Athena.
3. Use AWS Data Exchange for governed subscription and distribution of supported third-party data products.
4. Use Amazon OpenSearch Service.
5. Use Amazon Quick for managed business intelligence and visualization where it meets the dashboard requirements.

Correct Answer: C

 

Correct Answer

Answer C is correct because Data Exchange is designed to discover, subscribe to, and use third-party data products rather than building a custom data marketplace from scratch. This directly meets the decisive requirement: AWS Data Exchange.

Incorrect Answers

Answer A is incorrect because Redshift is a managed analytical warehouse designed for complex SQL across large curated datasets. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Data Exchange.

Answer B is incorrect because Athena provides serverless SQL queries over data in S3 and is well suited to ad-hoc analytical access without managing database infrastructure. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Data Exchange.

Answer D is incorrect because OpenSearch is suited to indexed full-text search, log analytics, and dashboard-oriented exploration. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Data Exchange.

Answer E is incorrect because The current SAA-C03 scope lists Amazon Quick as an analytics and visualization service; it is intended for managed BI dashboards. It may be useful elsewhere, but it does not meet the decisive requirement here: AWS Data Exchange.

 

Question 31

A delivery stream writes many tiny S3 objects every few seconds and downstream query performance suffers. Which ingestion parameter should be tuned?

1. Scale or optimize the stream consumers.
2. Increase delivery buffering size or interval while staying inside the allowed freshness requirement.
3. Partition the analytical dataset by frequently filtered attributes such as date.
4. Use a partition-key strategy that distributes records more evenly across shards.
5. Convert analytical data to a compressed columnar format such as Parquet.

Correct Answer: B

 

Correct Answer

Answer B is correct because Larger buffers reduce the number of tiny delivered objects and can improve downstream storage/query efficiency. This directly meets the decisive requirement: increase buffer size/interval within latency requirement.

Incorrect Answers

Answer A is incorrect because If producer/stream capacity is sufficient but iterator age continues to grow, the consumer side is not keeping up. It may be useful elsewhere, but it does not meet the decisive requirement here: increase buffer size/interval within latency requirement.

Answer C is incorrect because Partition pruning lets query engines avoid scanning unrelated object prefixes/partitions. It may be useful elsewhere, but it does not meet the decisive requirement here: increase buffer size/interval within latency requirement.

Answer D is incorrect because Even key distribution avoids concentrating most traffic on one shard. It may be useful elsewhere, but it does not meet the decisive requirement here: increase buffer size/interval within latency requirement.

Answer E is incorrect because Columnar compressed formats reduce bytes read for analytics that access only selected columns. It may be useful elsewhere, but it does not meet the decisive requirement here: increase buffer size/interval within latency requirement.

 

Question 32

A Kinesis stream has one hot shard because all devices use the same partition key. Which change improves throughput distribution?

1. Increase delivery buffering size or interval while staying inside the allowed freshness requirement.
2. Partition the analytical dataset by frequently filtered attributes such as date.
3. Convert analytical data to a compressed columnar format such as Parquet.
4. Use a partition-key strategy that distributes records more evenly across shards.
5. Scale or optimize the stream consumers.

Correct Answer: D

 

Correct Answer

Answer D is correct because Even key distribution avoids concentrating most traffic on one shard. This directly meets the decisive requirement: better partition-key distribution.

Incorrect Answers

Answer A is incorrect because Larger buffers reduce the number of tiny delivered objects and can improve downstream storage/query efficiency. It may be useful elsewhere, but it does not meet the decisive requirement here: better partition-key distribution.

Answer B is incorrect because Partition pruning lets query engines avoid scanning unrelated object prefixes/partitions. It may be useful elsewhere, but it does not meet the decisive requirement here: better partition-key distribution.

Answer C is incorrect because Columnar compressed formats reduce bytes read for analytics that access only selected columns. It may be useful elsewhere, but it does not meet the decisive requirement here: better partition-key distribution.

Answer E is incorrect because If producer/stream capacity is sufficient but iterator age continues to grow, the consumer side is not keeping up. It may be useful elsewhere, but it does not meet the decisive requirement here: better partition-key distribution.

 

Question 33

Athena scans far more bytes than expected because one large table is not partitioned by date. Which data-layout change can reduce scan work?

1. Scale or optimize the stream consumers.
2. Increase delivery buffering size or interval while staying inside the allowed freshness requirement.
3. Use a partition-key strategy that distributes records more evenly across shards.
4. Partition the analytical dataset by frequently filtered attributes such as date.
5. Convert analytical data to a compressed columnar format such as Parquet.

Correct Answer: D

 

Correct Answer

Answer D is correct because Partition pruning lets query engines avoid scanning unrelated object prefixes/partitions. This directly meets the decisive requirement: partition by common filter key.

Incorrect Answers

Answer A is incorrect because If producer/stream capacity is sufficient but iterator age continues to grow, the consumer side is not keeping up. It may be useful elsewhere, but it does not meet the decisive requirement here: partition by common filter key.

Answer B is incorrect because Larger buffers reduce the number of tiny delivered objects and can improve downstream storage/query efficiency. It may be useful elsewhere, but it does not meet the decisive requirement here: partition by common filter key.

Answer C is incorrect because Even key distribution avoids concentrating most traffic on one shard. It may be useful elsewhere, but it does not meet the decisive requirement here: partition by common filter key.

Answer E is incorrect because Columnar compressed formats reduce bytes read for analytics that access only selected columns. It may be useful elsewhere, but it does not meet the decisive requirement here: partition by common filter key.

 

Question 34

Analytics queries repeatedly read verbose CSV and are I/O-bound. Which storage format can improve scan efficiency?

1. Use a partition-key strategy that distributes records more evenly across shards.
2. Partition the analytical dataset by frequently filtered attributes such as date.
3. Convert analytical data to a compressed columnar format such as Parquet.
4. Scale or optimize the stream consumers.
5. Increase delivery buffering size or interval while staying inside the allowed freshness requirement.

Correct Answer: C

 

Correct Answer

Answer C is correct because Columnar compressed formats reduce bytes read for analytics that access only selected columns. This directly meets the decisive requirement: compressed columnar format such as Parquet.

Incorrect Answers

Answer A is incorrect because Even key distribution avoids concentrating most traffic on one shard. It may be useful elsewhere, but it does not meet the decisive requirement here: compressed columnar format such as Parquet.

Answer B is incorrect because Partition pruning lets query engines avoid scanning unrelated object prefixes/partitions. It may be useful elsewhere, but it does not meet the decisive requirement here: compressed columnar format such as Parquet.

Answer D is incorrect because If producer/stream capacity is sufficient but iterator age continues to grow, the consumer side is not keeping up. It may be useful elsewhere, but it does not meet the decisive requirement here: compressed columnar format such as Parquet.

Answer E is incorrect because Larger buffers reduce the number of tiny delivered objects and can improve downstream storage/query efficiency. It may be useful elsewhere, but it does not meet the decisive requirement here: compressed columnar format such as Parquet.

 

Question 35

Kinesis iterator age rises continuously even after stream capacity is increased. Which component should be scaled next?

1. Scale or optimize the stream consumers.
2. Partition the analytical dataset by frequently filtered attributes such as date.
3. Use a partition-key strategy that distributes records more evenly across shards.
4. Convert analytical data to a compressed columnar format such as Parquet.
5. Increase delivery buffering size or interval while staying inside the allowed freshness requirement.

Correct Answer: A

 

Correct Answer

Answer A is correct because If producer/stream capacity is sufficient but iterator age continues to grow, the consumer side is not keeping up. This directly meets the decisive requirement: consumer processing capacity.

Incorrect Answers

Answer B is incorrect because Partition pruning lets query engines avoid scanning unrelated object prefixes/partitions. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer C is incorrect because Even key distribution avoids concentrating most traffic on one shard. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer D is incorrect because Columnar compressed formats reduce bytes read for analytics that access only selected columns. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.

Answer E is incorrect because Larger buffers reduce the number of tiny delivered objects and can improve downstream storage/query efficiency. It may be useful elsewhere, but it does not meet the decisive requirement here: consumer processing capacity.