Amazon AWS Certified Data Engineer - Associate DEA-C01 Practice Test 17
Question 160
A recovery rerun mistakenly skips an object because bookmark state advanced before output was validated. Which evidence would BEST verify that the intended pipeline outcome occurred?
- The relevant validation evidence is a renamed resource label that does not show an authorization or processing result.
- The relevant validation evidence is job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
- The relevant validation evidence is a successful ping to an unrelated service that did not process the data.
- The relevant validation evidence is an old architecture drawing without the current query plan or job-run identifier.
Explanation
Correct Answer
B:Glue bookmarks track supported source progress across runs but do not replace destination commit auditing.
Incorrect Answers
A:A renamed resource label that does not show an authorization or processing result. This is not sufficient evidence for Glue job bookmarks for incremental S3 reads. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition. Glue bookmarks track supported source progress across runs but do not replace destination commit auditing.
C:A successful ping to an unrelated service that did not process the data. This is not sufficient evidence for Glue job bookmarks for incremental S3 reads. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition. Glue bookmarks track supported source progress across runs but do not replace destination commit auditing.
D:An old architecture drawing without the current query plan or job-run identifier. This is not sufficient evidence for Glue job bookmarks for incremental S3 reads. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition. Glue bookmarks track supported source progress across runs but do not replace destination commit auditing.
Question 161
A recovery rerun mistakenly skips an object because bookmark state advanced before output was validated. Which underlying AWS service principle BEST explains the recommended approach?
- The governing design principle is that a successful IAM login means every analytics dataset is readable and correctly transformed.
- The governing design principle is that increasing a resource name length automatically increases AWS service throughput.
- The governing design principle is that glue bookmarks track supported source progress across runs but do not replace destination commit auditing.
- The governing design principle is that a pipeline schedule guarantees that every downstream source record was processed correctly.
Explanation
Correct Answer
C:Implement by choosing to configure and explicitly manage Glue job bookmarks and source transformation context, with controlled reset or rewind. Verify with job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Incorrect Answers
A:A successful IAM login means every analytics dataset is readable and correctly transformed. That generalization conflicts with Glue job bookmarks for incremental S3 reads. Instead glue bookmarks track supported source progress across runs but do not replace destination commit auditing. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
B:Increasing a resource name length automatically increases AWS service throughput. That generalization conflicts with Glue job bookmarks for incremental S3 reads. Instead glue bookmarks track supported source progress across runs but do not replace destination commit auditing. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
D:A pipeline schedule guarantees that every downstream source record was processed correctly. That generalization conflicts with Glue job bookmarks for incremental S3 reads. Instead glue bookmarks track supported source progress across runs but do not replace destination commit auditing. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Question 162
A recovery rerun mistakenly skips an object because bookmark state advanced before output was validated. An incorrect proposal is: Rename every source file so the bookmark appears empty on each run. What material risk would follow from the proposed incorrect shortcut?
- The concrete failure is that a correctly encrypted S3 test object was the sole cause of an unrelated ISP outage.
- The concrete failure is that a harmless resource display-name edit caused all prior ETL jobs to lose their source schemas.
- The concrete failure is that reading a healthy CloudWatch log automatically erased every earlier record.
- The concrete failure is that assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced.
Explanation
Correct Answer
D:This is why the incorrect shortcut does not meet Glue job bookmarks for incremental S3 reads. The safer implementation is to configure and explicitly manage Glue job bookmarks and source transformation context, with controlled reset or rewind.
Incorrect Answers
A:A correctly encrypted S3 test object was the sole cause of an unrelated ISP outage. That unrelated assertion does not explain the actual Glue job bookmarks for incremental S3 reads risk. The relevant failure is that assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
B:A harmless resource display-name edit caused all prior ETL jobs to lose their source schemas. That unrelated assertion does not explain the actual Glue job bookmarks for incremental S3 reads risk. The relevant failure is that assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
C:Reading a healthy CloudWatch log automatically erased every earlier record. That unrelated assertion does not explain the actual Glue job bookmarks for incremental S3 reads risk. The relevant failure is that assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Check job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Question 163
A recovery rerun mistakenly skips an object because bookmark state advanced before output was validated. The observed failure is: Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Which recovery action BEST restores the required data-processing state?
- The supported recovery step is to assume crawler metadata alone stores a job's read checkpoint.
- The supported recovery step is to recover from a recorded bookmark point and reconcile output before committing continued incremental processing.
- The supported recovery step is to rename every source file so the bookmark appears empty on each run.
- The supported recovery step is to delete all target partitions on every incremental execution regardless of input changes.
Explanation
Correct Answer
B:This addresses the observed Glue job bookmarks for incremental S3 reads failure while preserving the required contract. Recheck job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Incorrect Answers
A:Assume crawler metadata alone stores a job's read checkpoint. Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Instead recover from a recorded bookmark point and reconcile output before committing continued incremental processing.
C:Rename every source file so the bookmark appears empty on each run. Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Instead recover from a recorded bookmark point and reconcile output before committing continued incremental processing.
D:Delete all target partitions on every incremental execution regardless of input changes. Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Instead recover from a recorded bookmark point and reconcile output before committing continued incremental processing.
Question 164
A recovery rerun mistakenly skips an object because bookmark state advanced before output was validated. The proposed implementation needs explicit operational validation. What combined implementation and validation decision offers the strongest evidence-based trade-off?
- The reviewed plan is to configure and explicitly manage Glue job bookmarks and source transformation context, with controlled reset or rewind; validate with job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
- The reviewed plan is to rename every source file so the bookmark appears empty on each run; treat a resource name as sufficient evidence.
- The reviewed plan is to delete all target partitions on every incremental execution regardless of input changes; omit the source-to-target verification.
- The reviewed plan is to assume crawler metadata alone stores a job's read checkpoint; do not check the failure conditions.
Explanation
Correct Answer
A:Glue bookmarks track supported source progress across runs but do not replace destination commit auditing. The change-review requirement is satisfied only when evidence is checked: Job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Incorrect Answers
B:Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. A resource label is not proof of Glue job bookmarks for incremental S3 reads.
C:Delete all target partitions on every incremental execution regardless of input changes. Omitting output verification can conceal failure. Glue bookmarks track supported source progress across runs but do not replace destination commit auditing. The accepted action is to configure and explicitly manage Glue job bookmarks and source transformation context, with controlled reset or rewind.
D:Assuming bookmarks guarantee atomic target writes can leave partial data even though input progress advanced. Verify with job bookmark enabled state, source path, previous run state, processed object IDs and target partition.
Question 165
A Glue JDBC read of a very large table uses one connection and exceeds the nightly ingestion window. Which AWS design or configuration BEST satisfies the stated processing requirement?
- Assume dividing by a low-cardinality flag creates equal JDBC read partitions.
- Increase Spark executors without checking JDBC source connections.
- Choose supported parallel JDBC partition bounds and fetch sizing that respect source database capacity.
- Use a Glue crawler as the database query load balancer.
Explanation
Correct Answer
C:Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. This directly addresses JDBC parallel reads and predicate partitioning in Glue.
Incorrect Answers
A:This does not meet JDBC parallel reads and predicate partitioning in Glue. Setting maximum parallelism without source-side limits can overload the transaction database.
B:Increase Spark executors without checking JDBC source connections. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Inspect jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
D:Setting maximum parallelism without source-side limits can overload the transaction database. The appropriate action is to choose supported parallel JDBC partition bounds and fetch sizing that respect source database capacity.
Question 166
A Glue JDBC read of a very large table uses one connection and exceeds the nightly ingestion window. Which evidence would BEST verify that the intended pipeline outcome occurred?
- A renamed resource label that does not show an authorization or processing result.
- A browser screenshot from another AWS account and time period.
- An old architecture drawing without the current query plan or job-run identifier.
- JDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
Explanation
Correct Answer
D:Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. These observations directly test JDBC parallel reads and predicate partitioning in Glue.
Incorrect Answers
A:A renamed resource label that does not show an authorization or processing result. This is not sufficient evidence for JDBC parallel reads and predicate partitioning in Glue. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
B:A browser screenshot from another AWS account and time period. This is not sufficient evidence for JDBC parallel reads and predicate partitioning in Glue. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
C:An old architecture drawing without the current query plan or job-run identifier. This is not sufficient evidence for JDBC parallel reads and predicate partitioning in Glue. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
Question 167
A Glue JDBC read of a very large table uses one connection and exceeds the nightly ingestion window. Which underlying AWS service principle BEST explains the recommended approach?
- Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
- A CloudWatch metric label is equivalent to a successful target table data-quality validation.
- Increasing a resource name length automatically increases AWS service throughput.
- A pipeline schedule guarantees that every downstream source record was processed correctly.
Explanation
Correct Answer
A:This is the relevant technical mechanism for JDBC parallel reads and predicate partitioning in Glue. Implement by choosing to choose supported parallel JDBC partition bounds and fetch sizing that respect source database capacity.
Incorrect Answers
B:A CloudWatch metric label is equivalent to a successful target table data-quality validation. That generalization conflicts with JDBC parallel reads and predicate partitioning in Glue. Instead parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
C:Increasing a resource name length automatically increases AWS service throughput. That generalization conflicts with JDBC parallel reads and predicate partitioning in Glue. Instead parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
D:A pipeline schedule guarantees that every downstream source record was processed correctly. That generalization conflicts with JDBC parallel reads and predicate partitioning in Glue. Instead parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
Question 168
A Glue JDBC read of a very large table uses one connection and exceeds the nightly ingestion window. An incorrect proposal is: Assume dividing by a low-cardinality flag creates equal JDBC read partitions. What material risk would follow from the proposed incorrect shortcut?
- A correctly encrypted S3 test object was the sole cause of an unrelated ISP outage.
- Setting maximum parallelism without source-side limits can overload the transaction database.
- A harmless resource display-name edit caused all prior ETL jobs to lose their source schemas.
- The presence of an accurate job timestamp automatically created unauthorized database credentials.
Explanation
Correct Answer
B:This is why the incorrect shortcut does not meet JDBC parallel reads and predicate partitioning in Glue. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
Incorrect Answers
A:A correctly encrypted S3 test object was the sole cause of an unrelated ISP outage. That unrelated assertion does not explain the actual JDBC parallel reads and predicate partitioning in Glue risk. The relevant failure is that setting maximum parallelism without source-side limits can overload the transaction database. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
C:A harmless resource display-name edit caused all prior ETL jobs to lose their source schemas. That unrelated assertion does not explain the actual JDBC parallel reads and predicate partitioning in Glue risk. The relevant failure is that setting maximum parallelism without source-side limits can overload the transaction database. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
D:The presence of an accurate job timestamp automatically created unauthorized database credentials. That unrelated assertion does not explain the actual JDBC parallel reads and predicate partitioning in Glue risk. The relevant failure is that setting maximum parallelism without source-side limits can overload the transaction database. Check jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
Question 169
A Glue JDBC read of a very large table uses one connection and exceeds the nightly ingestion window. The observed failure is: Setting maximum parallelism without source-side limits can overload the transaction database. Which recovery action BEST restores the required data-processing state?
- Increase Spark executors without checking JDBC source connections.
- Use a Glue crawler as the database query load balancer.
- Assume dividing by a low-cardinality flag creates equal JDBC read partitions.
- Tune partition bounds and connection count against measured source read pressure and reconcile row totals.
Explanation
Correct Answer
D:This addresses the observed JDBC parallel reads and predicate partitioning in Glue failure while preserving the required contract. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew.
Incorrect Answers
A:Increase Spark executors without checking JDBC source connections. Setting maximum parallelism without source-side limits can overload the transaction database. Instead tune partition bounds and connection count against measured source read pressure and reconcile row totals.
B:Use a Glue crawler as the database query load balancer. Setting maximum parallelism without source-side limits can overload the transaction database. Instead tune partition bounds and connection count against measured source read pressure and reconcile row totals.
C:Assume dividing by a low-cardinality flag creates equal JDBC read partitions. Setting maximum parallelism without source-side limits can overload the transaction database. Instead tune partition bounds and connection count against measured source read pressure and reconcile row totals.
Question 170
A database source sees excessive concurrent connections after indiscriminately increasing Glue read partitions. What is the MOST appropriate targeted implementation for this data-engineering situation?
- In this operational case, increase Spark executors without checking JDBC source connections.
- In this operational case, choose supported parallel JDBC partition bounds and fetch sizing that respect source database capacity.
- In this operational case, use a Glue crawler as the database query load balancer.
- In this operational case, assume dividing by a low-cardinality flag creates equal JDBC read partitions.
Explanation
Correct Answer
B:Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Validate with jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
Incorrect Answers
A:Increase Spark executors without checking JDBC source connections. Parallel JDBC extraction can increase throughput but generates concurrent source queries and must avoid skew. Inspect jDBC partition predicates, query plan, active DB sessions, read duration and sampled row counts.
C:Setting maximum parallelism without source-side limits can overload the transaction database. Use a Glue crawler as the database query load balancer.
D:This does not meet JDBC parallel reads and predicate partitioning in Glue. The supported design is to choose supported parallel JDBC partition bounds and fetch sizing that respect source database capacity.