Amazon AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 62
Question 610
A Kinesis consumer reads events too slowly and processing lag grows. What metric should direct scaling decisions?
- Monitor IteratorAge and processing throughput against stream capacity.
- Choose more evenly distributed Kinesis partition keys for hot data.
- Use Kinesis Data Streams or another durable ordered log for replay-based consumers.
- Use Amazon Data Firehose with an S3 destination and supported buffering.
Explanation
Correct Answer
A:Consumer lag measures how far behind event processing has fallen. Increase consumer capacity and compare IteratorAge.
Incorrect Answers
B:Partition key skew can saturate a shard independent of total stream capacity. Monitor IteratorAge and processing throughput against stream capacity. Increase consumer capacity and compare IteratorAge.
C:SNS pub/sub delivery does not replace a retention-based replayable data stream. Monitor IteratorAge and processing throughput against stream capacity. Increase consumer capacity and compare IteratorAge.
D:Firehose manages delivery of streaming records to supported destinations. Monitor IteratorAge and processing throughput against stream capacity. Increase consumer capacity and compare IteratorAge.
Question 611
A business needs streaming data delivered to a supported analytics destination with failed-record backups. What delivery configuration fits?
- Configure appropriate Firehose buffering size and interval.
- Use Firehose with an approved destination and S3 error backup settings.
- Use Kinesis Data Streams with sufficient configured retention.
- Use Amazon Data Firehose with an S3 destination and supported buffering.
Explanation
Correct Answer
B:Managed delivery pipelines can preserve failed transformed/delivered records for review. Force one transformation failure and inspect backup output.
Incorrect Answers
A:Delivery buffering batches records into larger objects under eligible conditions. Use Firehose with an approved destination and S3 error backup settings. Force one transformation failure and inspect backup output.
C:Data Streams supports multiple consumers with ordered shard records and retention. Managed delivery pipelines can preserve failed transformed/delivered records for review.
D:Firehose manages delivery of streaming records to supported destinations. Use Firehose with an approved destination and S3 error backup settings. Force one transformation failure and inspect backup output.
Question 612
A team expects an SNS topic to provide ordered persistent stream replay for days. What architecture is more appropriate?
- Use Kinesis Data Streams or another durable ordered log for replay-based consumers.
- Choose more evenly distributed Kinesis partition keys for hot data.
- Use Amazon Data Firehose with an S3 destination and supported buffering.
- Use Kinesis Data Streams with retention shorter than the longest outage and still promise multi-day replay.
Explanation
Correct Answer
A:SNS pub/sub delivery does not replace a retention-based replayable data stream. Restart a reader from a stored sequence number.
Incorrect Answers
B:Partition key skew can saturate a shard independent of total stream capacity. Use Kinesis Data Streams or another durable ordered log for replay-based consumers. Restart a reader from a stored sequence number.
C:Firehose manages delivery of streaming records to supported destinations. Use Kinesis Data Streams or another durable ordered log for replay-based consumers. Restart a reader from a stored sequence number.
D:An ordered stream is helpful only when its retention exceeds the lag or replay horizon. An insufficient Kinesis retention setting would age out records before consumers can replay them.
Question 613
A data lake needs scheduled serverless Spark jobs to clean and transform incoming files. Which managed ETL service fits?
- Use AWS Lake Formation permissions with Data Catalog integration.
- Enable appropriate Glue job bookmarks and idempotent target writes.
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Use AWS Glue Schema Registry for supported streaming serialization formats.
Explanation
Correct Answer
C:Glue provides managed data transformation on supported Spark runtimes. Run the ETL and validate output partitions.
Incorrect Answers
A:Lake Formation enforces supported fine-grained permissions on governed data lakes. Use AWS Glue ETL jobs with appropriate triggers and job bookmarks. Run the ETL and validate output partitions.
B:Bookmarks track supported sources' processed state but still need safe output handling. Use AWS Glue ETL jobs with appropriate triggers and job bookmarks. Run the ETL and validate output partitions.
D:Schema Registry supports compatible schema evolution and discovery. Use AWS Glue ETL jobs with appropriate triggers and job bookmarks. Run the ETL and validate output partitions.
Question 614
Files arrive with evolving schemas and analysts need centralized table metadata for Athena. Which service component fits?
- Refresh Glue Data Catalog partitions using crawler or suitable partition registration.
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Use AWS Lake Formation permissions with Data Catalog integration.
Explanation
Correct Answer
B:The Data Catalog stores table and partition metadata for AWS analytics services. Query a registered table in Athena.
Incorrect Answers
A:Catalog partitions determine which paths are visible to partition-aware queries. Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas. Query a registered table in Athena.
C:Glue provides managed data transformation on supported Spark runtimes. Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas. Query a registered table in Athena.
D:Lake Formation enforces supported fine-grained permissions on governed data lakes. Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas. Query a registered table in Athena.
Question 615
An incremental Glue ETL job repeatedly reprocesses old data and duplicates target records. What incremental control helps?
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Enable appropriate Glue job bookmarks and idempotent target writes.
- Use AWS Glue Schema Registry for supported streaming serialization formats.
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
Explanation
Correct Answer
B:Bookmarks track supported sources' processed state but still need safe output handling. Run the job twice and inspect duplicate counts.
Incorrect Answers
A:Glue provides managed data transformation on supported Spark runtimes. Enable appropriate Glue job bookmarks and idempotent target writes. Run the job twice and inspect duplicate counts.
C:Schema Registry supports compatible schema evolution and discovery. Enable appropriate Glue job bookmarks and idempotent target writes. Run the job twice and inspect duplicate counts.
D:The Data Catalog stores table and partition metadata for AWS analytics services. Bookmarks track supported sources' processed state but still need safe output handling.
Question 616
A table's newly created S3 partitions are missing in query results. What metadata must be updated?
- Use AWS Lake Formation permissions with Data Catalog integration.
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
- Refresh Glue Data Catalog partitions using crawler or suitable partition registration.
- Convert data to partitioned columnar Parquet format through an ETL job.
Explanation
Correct Answer
C:Catalog partitions determine which paths are visible to partition-aware queries. Query the newly registered partition.
Incorrect Answers
A:Lake Formation enforces supported fine-grained permissions on governed data lakes. Refresh Glue Data Catalog partitions using crawler or suitable partition registration. Query the newly registered partition.
B:The Data Catalog stores table and partition metadata for AWS analytics services. Refresh Glue Data Catalog partitions using crawler or suitable partition registration. Query the newly registered partition.
D:Columnar compression and partition pruning reduce read volume for analytics queries. Refresh Glue Data Catalog partitions using crawler or suitable partition registration. Query the newly registered partition.
Question 617
A regulated lake must control read access at table and column granularity across Athena and Glue. Which governance service fits?
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Use AWS Lake Formation permissions with Data Catalog integration.
- Use AWS Glue Schema Registry for supported streaming serialization formats.
Explanation
Correct Answer
C:Lake Formation enforces supported fine-grained permissions on governed data lakes. Query restricted columns as distinct roles.
Incorrect Answers
A:The Data Catalog stores table and partition metadata for AWS analytics services. Use AWS Lake Formation permissions with Data Catalog integration. Query restricted columns as distinct roles.
B:Glue provides managed data transformation on supported Spark runtimes. Use AWS Lake Formation permissions with Data Catalog integration. Query restricted columns as distinct roles.
D:Schema Registry supports compatible schema evolution and discovery. Use AWS Lake Formation permissions with Data Catalog integration. Query restricted columns as distinct roles.
Question 618
A Glue job runs out of memory with large skewed joins despite available executor CPU. What workload tuning matters?
- Review Spark partition skew, worker type and join strategy before merely adding CPUs.
- Refresh Glue Data Catalog partitions using crawler or suitable partition registration.
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Convert data to partitioned columnar Parquet format through an ETL job.
Explanation
Correct Answer
A:Distributed ETL performance depends on partition size and memory behavior. Inspect stage skew and out-of-memory events.
Incorrect Answers
B:Catalog partitions determine which paths are visible to partition-aware queries. Review Spark partition skew, worker type and join strategy before merely adding CPUs. Inspect stage skew and out-of-memory events.
C:Glue provides managed data transformation on supported Spark runtimes. Review Spark partition skew, worker type and join strategy before merely adding CPUs. Inspect stage skew and out-of-memory events.
D:Columnar compression and partition pruning reduce read volume for analytics queries. Review Spark partition skew, worker type and join strategy before merely adding CPUs. Inspect stage skew and out-of-memory events.
Question 619
A CSV archive is expensive to scan repeatedly in Athena. What Glue transformation may improve it?
- Convert data to partitioned columnar Parquet format through an ETL job.
- Refresh Glue Data Catalog partitions using crawler or suitable partition registration.
- Use AWS Glue ETL jobs with appropriate triggers and job bookmarks.
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
Explanation
Correct Answer
A:Columnar compression and partition pruning reduce read volume for analytics queries. Compare Athena bytes scanned after conversion.
Incorrect Answers
B:Catalog partitions determine which paths are visible to partition-aware queries. Convert data to partitioned columnar Parquet format through an ETL job. Compare Athena bytes scanned after conversion.
C:Glue provides managed data transformation on supported Spark runtimes. Convert data to partitioned columnar Parquet format through an ETL job. Compare Athena bytes scanned after conversion.
D:The Data Catalog stores table and partition metadata for AWS analytics services. Convert data to partitioned columnar Parquet format through an ETL job. Compare Athena bytes scanned after conversion.
Question 620
An application needs a schema registry to validate evolving event formats between producers and consumers. Which AWS integration is useful?
- Use AWS Glue Data Catalog with reviewed crawlers or explicit schemas.
- Use AWS Lake Formation permissions with Data Catalog integration.
- Enable appropriate Glue job bookmarks and idempotent target writes.
- Use AWS Glue Schema Registry for supported streaming serialization formats.
Explanation
Correct Answer
D:Schema Registry supports compatible schema evolution and discovery. Reject an incompatible test schema version.
Incorrect Answers
A:The Data Catalog stores table and partition metadata for AWS analytics services. Use AWS Glue Schema Registry for supported streaming serialization formats. Reject an incompatible test schema version.
B:Lake Formation enforces supported fine-grained permissions on governed data lakes. Use AWS Glue Schema Registry for supported streaming serialization formats. Reject an incompatible test schema version.
C:Bookmarks track supported sources' processed state but still need safe output handling. Schema Registry supports compatible schema evolution and discovery.