100% Latest & Updated Amazon AWS Certified Data Engineer - Associate DEA-C01 Practice Test Questions, Exam Dumps & Verified Answers!
30 Days Free Updates, Instant Download!
Question 30
A team applies a reviewed fix for this situation: A Glue data job decrypts source objects using a wildcard KMS permission spanning unrelated client keys. Which observed outcome would most convincingly validate the fix?
- A nominated team member can access only the project assets approved for the role.
- The production role reads an approved sample without expanded administrative rights.
- The partner accesses only intended assets and access ends when approval expires.
- The job decrypts authorized data but receives denial on a neighboring client's key.
Explanation
Correct Answer
D:The job decrypts authorized data but receives denial on a neighboring client's key. This verifies the result of scope KMS actions to intended keys and add appropriate conditions where supported.
Incorrect Answers
A:This would verify another situation: An admin provisions a SageMaker Unified Studio project but the intended team cannot open governed domain assets despite being added to a general account group. It does not validate the current case.
B:This would verify another situation: An S3 lake read works with an administrator role but fails with the production ETL role, and the team assumes the dataset is corrupt. It does not validate the current case.
C:This would verify another situation: A Redshift datashare grants partner analysts access to a customer data mart without a purpose or term limit. It does not validate the current case.
Question 31
A Redshift SQL view joins two tables on customer name, and a same-name collision merges unrelated accounts. Which technical dependency most directly explains this behavior?
- Completeness checks do not assess referential validity against current mappings.
- Presence checking does not enforce the field's required format contract.
- A display label is not a stable relational join key.
- Single-metric quality scoring hides unintended data loss.
Explanation
Correct Answer
C:A display label is not a stable relational join key. The corresponding fix is to join on documented unique customer identifiers and enforce suitable grain.
Incorrect Answers
A:This diagnosis addresses another AWS case: A Glue transformation produces 100% nonnull rows but an upstream category mapping is outdated and mislabels every new source code. The dependency here is different.
B:This diagnosis addresses another AWS case: A Glue data quality rule checks only that an email column is nonnull but accepts invalid malformed addresses. The dependency here is different.
D:This diagnosis addresses another AWS case: A data team declares a successful cleanup because duplicate percentage fell, but the job also dropped valid unique rows. The dependency here is different.
Question 32
After resolving the following incident, which ongoing safeguard most directly prevents recurrence? An AWS DMS migration creates a new database table but foreign-key dependencies fail during cutover because schema compatibility was never rehearsed.
- Document fact grain and validate cardinality in data model reviews.
- Add temporal join and duplicate-cardinality tests to dimensional models.
- Add schema compatibility checks before crawler runs.
- Maintain schema/constraint migration checklists and rollback plans.
Explanation
Correct Answer
D:Maintain schema/constraint migration checklists and rollback plans. This safeguards against a recurring cause: Data migration alone does not resolve source/target schema and constraint differences.
Incorrect Answers
A:That safeguard addresses another case: A Redshift star schema mixes order-line and order-header facts in the same table and reports double-count revenue. It is not the control for this recurring failure.
B:That safeguard addresses another case: A warehouse fact table is joined to customer history using only customer ID, multiplying matches across changing customer versions. It is not the control for this recurring failure.
C:That safeguard addresses another case: A Glue crawler discovers an S3 path containing two incompatible CSV header layouts and produces a wide schema with ambiguous columns. It is not the control for this recurring failure.
Question 33
A marketing report uses COUNT(*) of event records as its active user metric and overstates adoption when users retry. Which technical dependency most directly explains this behavior?
- The workflow's success criteria omit an operationally required delivery dependency.
- The data quality contract omits numeric precision and tolerance.
- Events and distinct users are different measurement grains.
- Failure evidence is lost with ephemeral compute nodes.
Explanation
Correct Answer
C:Events and distinct users are different measurement grains. The corresponding fix is to aggregate by validated stable user identity and the intended time-window rules.
Incorrect Answers
A:This diagnosis addresses another AWS case: A Step Functions run is marked succeeded although a downstream notification publish fails and no incident ticket is opened. The dependency here is different.
B:This diagnosis addresses another AWS case: An upstream system changes a decimal field from two to four digits of precision and rounding introduces revenue drift. The dependency here is different.
D:This diagnosis addresses another AWS case: An EMR job intermittently fails due to capacity interruptions but its logs are stored only on the terminated instance. The dependency here is different.
Question 34
A team applies a reviewed fix for this situation: A DynamoDB table uses the same partition key for all new daily events, producing throttled writes during each morning burst. Which observed outcome would most convincingly validate the fix?
- An overwritten fixture can be restored within the intended recovery interval.
- An injected failed publish leaves the last approved snapshot accessible.
- Peak writes distribute without losing required entity queries.
- Transaction lookups are efficient without exposing cross-tenant records.
Explanation
Correct Answer
C:Peak writes distribute without losing required entity queries. This verifies the result of redesign keys using appropriately distributed entity/shard identifiers while preserving retrieval needs.
Incorrect Answers
A:This would verify another situation: An S3 lifecycle rule expires noncurrent object versions immediately, preventing recovery from accidental overwrites. It does not validate the current case.
B:This would verify another situation: A nightly lake job publishes versioned S3 snapshots but deletes all old snapshots before the replacement is fully validated. It does not validate the current case.
D:This would verify another situation: A DynamoDB table uses composite key `tenant#date` but most reads are by `transaction_id` only and become scans. It does not validate the current case.
Question 35
A data privacy analyst needs to identify PII in unknown S3 datasets but assumes object names reveal sensitive content. Select TWO complementary measures that correctly repair and validate this failure.
- Review the required access-point and identity/bucket permissions for the approved path.
- Controlled sensitive fixtures are identified and non-sensitive examples are handled appropriately.
- Use approved Macie sensitive-data discovery and validated data classifications with access review.
- The crawler accesses the RDS schema from its actual VPC context.
- Classify every filename containing `customer` as automatically harmless.
Explanation
Correct Answers
B:Controlled sensitive fixtures are identified and non-sensitive examples are handled appropriately. This is the observable result required before accepting the correction.
C:Use approved Macie sensitive-data discovery and validated data classifications with access review. The underlying cause is: Object naming is not a reliable classifier of data content.
Incorrect Answers
A:That corrective action is for a different case: An S3 lake access point exists but a cross-account principal still receives AccessDenied when its IAM and bucket policies are not aligned. It does not resolve this incident.
D:This would validate a different case: A Glue crawler in a VPC fails connection to a private RDS source even with the right IAM service role. It does not verify the outcome required here.
E:Classify every filename containing `customer` as automatically harmless. That shortcut conflicts with the stated requirement.
Question 36
A team applies a reviewed fix for this situation: An S3 event notification triggers Glue for every file of a large export, producing hundreds of competing jobs. Which observed outcome would most convincingly validate the fix?
- Peak concurrent calls stay bounded and all tasks finish without escalating throttles.
- A failed release can restore the prior script and reproduce its output.
- One controlled export yields the intended bounded job count and all objects are processed.
- A simulated transient failure recovers and persistent failures produce timely alerts.
Explanation
Correct Answer
C:One controlled export yields the intended bounded job count and all objects are processed. This verifies the result of aggregate object events or publish a manifest/ready signal and start bounded ETL batches.
Incorrect Answers
A:This would verify another situation: A pipeline launches thousands of concurrent Step Functions Map tasks and saturates a small partner API. It does not validate the current case.
B:This would verify another situation: A data engineering team manually changes Glue scripts on S3 without source control or a rollback identifier. It does not validate the current case.
D:This would verify another situation: An AWS MWAA DAG has no task-level retries or SLA alerts, and a brief Glue API throttle causes a missed daily close. It does not validate the current case.
Question 37
A warehouse fact table is joined to customer history using only customer ID, multiplying matches across changing customer versions. Select THREE complementary measures that correctly repair and validate, and prevent recurrence of this failure.
- Join using correct surrogate keys or as-of effective dates for SCD history.
- A test order maps to exactly the customer attributes valid at transaction time.
- Add temporal join and duplicate-cardinality tests to dimensional models.
- Use supported S3 Object Lock legal hold/retention and review lifecycle interactions.
- Delete old dimension versions to force a unique join.
- Expired sessions are denied immediately even if the physical item persists briefly.
Explanation
Correct Answers
A:Join using correct surrogate keys or as-of effective dates for SCD history. The underlying cause is: Dimension joins ignore the effective time range for slowly changing records.
B:A test order maps to exactly the customer attributes valid at transaction time. This is the observable result required before accepting the correction.
C:Add temporal join and duplicate-cardinality tests to dimensional models. This protects against the following risk: Dimension joins ignore the effective time range for slowly changing records.
Incorrect Answers
D:That corrective action is for a different case: A legal-hold dataset is subject to an S3 Lifecycle expiration rule that would remove evidence at the usual age. It does not resolve this incident.
E:Delete old dimension versions to force a unique join. That shortcut conflicts with the stated requirement.
F:This would validate a different case: A DynamoDB table contains session records with expired timestamps but items remain visible immediately after expiry. It does not verify the outcome required here.
Question 38
A table claims a `region` partition key but the written S3 paths use `geography=...`, so Athena pruning fails. Select THREE complementary measures that correctly repair and validate, and prevent recurrence of this failure.
- Rename random S3 objects without updating the manifest or catalog.
- Align approved partition key definitions and S3 path generation or repair registered metadata.
- Contract-test the partition naming convention at writer and catalog boundaries.
- Region-filtered queries prune to the expected physical prefixes.
- Reassess partition keys/granularity and apply efficient columnar file layout.
- A controlled subject deletion is absent from permitted current and derived query paths.
Explanation
Correct Answers
B:Align approved partition key definitions and S3 path generation or repair registered metadata. The underlying cause is: Partition metadata and physical prefix naming do not agree.
C:Contract-test the partition naming convention at writer and catalog boundaries. This protects against the following risk: Partition metadata and physical prefix naming do not agree.
D:Region-filtered queries prune to the expected physical prefixes. This is the observable result required before accepting the correction.
Incorrect Answers
A:Rename random S3 objects without updating the manifest or catalog. That shortcut conflicts with the stated requirement.
E:That corrective action is for a different case: An Athena table uses daily partitions on a timestamp with millions of near-empty partition folders. It does not resolve this incident.
F:This would validate a different case: An analytics partition must be removed under a deletion request, yet its historical copies persist in untracked S3 archive paths. It does not verify the outcome required here.
Question 39
A daily S3 data lake output is present but the last successful ETL timestamp is two days old. Which technical dependency most directly explains this behavior?
- The reconciliation rule ignores documented transformation semantics and accepted rejects.
- Presence checking does not enforce the field's required format contract.
- The sample is too small or unrepresentative for a claimed whole-population guarantee.
- Object existence alone does not prove the current scheduled partition is fresh.
Explanation
Correct Answer
D:Object existence alone does not prove the current scheduled partition is fresh. The corresponding fix is to track a per-partition completion watermark and validate expected event-time coverage.
Incorrect Answers
A:This diagnosis addresses another AWS case: A data quality gate compares raw and target record totals without accounting for intentional filter exclusions, falsely failing clean batches. The dependency here is different.
B:This diagnosis addresses another AWS case: A Glue data quality rule checks only that an email column is nonnull but accepts invalid malformed addresses. The dependency here is different.
C:This diagnosis addresses another AWS case: A streaming pipeline evaluates quality from a random sample of 10 messages and assumes it proves a million-row batch is correct. The dependency here is different.
Question 40
A data team declares a successful cleanup because duplicate percentage fell, but the job also dropped valid unique rows. Select THREE complementary measures that correctly repair and validate, and prevent recurrence of this failure.
- Keep identifiers typed as strings and apply field-specific whitespace rules only.
- A shuffled fixture returns the same expected seven-day windows.
- Prefer fewer rows as sufficient proof of higher data quality.
- Accepted distinct keys remain complete while true duplicates are controlled.
- Compare uniqueness, completeness, reconciliation, and distribution against known fixtures.
- Require multidimensional quality gates before final publish.
Explanation
Correct Answers
D:Accepted distinct keys remain complete while true duplicates are controlled. This is the observable result required before accepting the correction.
E:Compare uniqueness, completeness, reconciliation, and distribution against known fixtures. The underlying cause is: Single-metric quality scoring hides unintended data loss.
F:Require multidimensional quality gates before final publish. This protects against the following risk: Single-metric quality scoring hides unintended data loss.
Incorrect Answers
A:That corrective action is for a different case: An analyst creates a Glue DataBrew visual cleaning step that trims all leading zeros from product IDs, breaking lookups. It does not resolve this incident.
B:This would validate a different case: A data science notebook uses a rolling 7-day average without ordering by event date, so results change with input row order. It does not verify the outcome required here.
C:Prefer fewer rows as sufficient proof of higher data quality. That shortcut conflicts with the stated requirement.
AWS Certified Data Engineer - Associate DEA-C01 Practice Tests