Question 300

Executives want each region on one row and each product class as a separate total column. Which transformation pattern fits?

  1. Redshift SQL window function
  2. SQL PIVOT transformation
  3. Athena SQL view
  4. SQL GROUP BY aggregate

Explanation

Correct Answer

B:

Pivoting turns distinct category values into separate columns for cross-tab reports in supported SQL engines; in this scenario, pivot category groups or use equivalent conditional aggregation according to the SQL dialect and expected categories.

Incorrect Answers

A:

Window functions calculate values across related rows while preserving row-level output; the current need is for presenting monthly category totals as columns, not computing a rolling average for each device.

C:

An Athena view stores a query definition for reuse over underlying tables; this helps with exposing a stable SQL projection over lake data, not presenting monthly category totals as columns.

D:

GROUP BY collapses input rows into groups and computes aggregate measures per distinct key; although useful for producing daily totals per store, it leaves the need for presenting monthly category totals as columns unmet.

Question 301

An asynchronously invoked enrichment function exhausts its retry attempts and failed payloads need investigation. Which feature captures the outcome?

  1. Reserved concurrency
  2. Provisioned concurrency
  3. Lambda memory/CPU setting
  4. Lambda asynchronous invocation destinations

Explanation

Correct Answer

D:

Lambda asynchronous destinations can send execution success or failure records to supported targets; in this scenario, configure an on-failure destination with appropriate permissions and monitor undelivered outcomes so failed events can be reprocessed.

Incorrect Answers

A:

Reserved concurrency allocates a concurrency limit and reservation to a specific Lambda function; this can support protecting a downstream database from too many simultaneous function calls, but would not address the need for recording failed asynchronous enrichments for repair.

B:

Provisioned concurrency pre-initializes a configured number of Lambda execution environments to reduce cold-start delay; this helps with stabilizing low-latency synchronous API responses, not recording failed asynchronous enrichments for repair.

C:

Lambda allocates more CPU capacity as the configured memory size increases; although useful for speeding CPU-intensive parsing in a Lambda transform, it leaves the need for recording failed asynchronous enrichments for repair unmet.

Question 302

A crawler unexpectedly changes a column type, and engineers must compare the previous catalog definition. What metadata capability helps?

  1. Data Catalog table version history
  2. Glue partition metadata
  3. Athena named database
  4. Glue Data Catalog table

Explanation

Correct Answer

A:

Data Catalog versions can preserve prior table metadata definitions for certain schema-management workflows; in this scenario, examine catalog table versions and record controlled schema changes rather than assuming S3 objects themselves were modified.

Incorrect Answers

B:

Catalog partition definitions map partition values to locations so compatible engines can prune object sets; although useful for registering new date partitions before reading them, it leaves the need for inspecting a prior table schema after a crawler update unmet.

C:

An Athena database is a logical namespace for tables in the underlying catalog; it would target organizing named datasets for a SQL workgroup, while the question is about inspecting a prior table schema after a crawler update.

D:

A Data Catalog table describes a dataset's location, columns and serialization for compatible query engines; its benefit is sharing an S3 table definition across Athena and Glue; that is different from inspecting a prior table schema after a crawler update.

Question 303

A one-minute Glue error metric briefly spikes during healthy retries and triggers noisy pages. Which alarm parameter can be tuned?

  1. CloudWatch alarm composite
  2. CloudWatch alarm evaluation periods
  3. SNS subscription filtering
  4. EventBridge failed-event DLQ

Explanation

Correct Answer

B:

Alarm evaluation periods govern how many datapoints must breach a threshold before the alarm changes state; in this scenario, require sustained breaching datapoints appropriate to the recovery objective, rather than alarming on one transient sample.

Incorrect Answers

A:

Composite alarms combine the states of other alarms using logical expressions; this helps with paging only when error rate and backlog are both unhealthy, not suppressing alerts for harmless single-sample spikes.

C:

SNS subscription filter policies select which topic messages are delivered to specific subscribers; it would target routing alerts by data pipeline ownership, while the question is about suppressing alerts for harmless single-sample spikes.

D:

A rule target can use an SQS dead-letter queue to retain undelivered events after retry exhaustion; its benefit is capturing notifications that fail delivery to the target; that is different from suppressing alerts for harmless single-sample spikes.

Question 304

A Lambda-based file parser loads a huge file into RAM and is killed before any results are produced. What programming approach helps?

  1. Set membership lookup
  2. Vectorized dataframe operation
  3. SQL predicate pushdown
  4. Chunked iterator processing

Explanation

Correct Answer

D:

Chunking processes large inputs in bounded pieces rather than loading an entire dataset into memory; in this scenario, stream or split input into bounded chunks and write incremental results, within Lambda runtime and storage constraints.

Incorrect Answers

A:

Hash-based set membership generally offers efficient repeated presence checks in memory; although useful for accelerating millions of repeated key lookups, it leaves the need for reading multi-gigabyte files within tight memory limits unmet.

B:

Vectorized data operations process columns or batches without per-row interpreter overhead; it would target reducing per-row Python transformation overhead, while the question is about reading multi-gigabyte files within tight memory limits.

C:

Predicate pushdown filters eligible records near the source and reduces data moved for ETL; its benefit is avoiding needless rows in a JDBC extraction; that is different from reading multi-gigabyte files within tight memory limits.

Question 305

A central analytics job must access a source account without distributing long-lived access keys. What identity model fits?

  1. Secrets Manager scheduled rotation
  2. Lambda execution IAM role
  3. Glue job service role
  4. STS AssumeRole for cross-account processing

Explanation

Correct Answer

D:

STS AssumeRole returns temporary credentials for an authorized role trust relationship; in this scenario, establish cross-account role trust and scoped permissions, then let the job assume the source role temporarily.

Incorrect Answers

A:

Secrets Manager rotation workflows can update supported credentials and associated secret versions; this can support rotating a relational password without embedding updates in code, but would not address the need for allowing an ETL account to read a partner account's approved objects.

B:

Lambda assumes its execution role to call permitted AWS APIs during function invocation; this helps with granting Lambda controlled read access to one S3 prefix, not allowing an ETL account to read a partner account's approved objects.

C:

A Glue job runs with an IAM role that authorizes its data, catalog and logging operations; although useful for enabling Glue to access controlled catalog and S3 resources, it leaves the need for allowing an ETL account to read a partner account's approved objects unmet.

Question 306

The platform already has an EventBridge scheduled rule for a weekly window and a Step Functions state machine with retries and branches. Which integration attaches the existing rule to the workflow?

  1. EventBridge Scheduler
  2. Step Functions with EventBridge rule
  3. Glue job trigger schedule
  4. Amazon MWAA DAG schedule

Explanation

Correct Answer

B:

An EventBridge scheduled rule can initiate a Step Functions state machine execution; in this scenario, route a scheduled EventBridge event to StartExecution while preserving orchestration logic in the state machine.

Incorrect Answers

A:

EventBridge Scheduler invokes supported targets on configured schedules with time-zone options; the current need is for periodically invoking a multi-step serverless data workflow, not launching a daily transform at a specified local time.

C:

Glue scheduled triggers start designated Glue jobs according to a configured schedule; this helps with running a single Glue ETL job on a fixed cadence, not periodically invoking a multi-step serverless data workflow.

D:

An Airflow DAG schedule defines recurring execution of an orchestration graph in Amazon MWAA; although useful for scheduling a dependency-aware Airflow workflow, it leaves the need for periodically invoking a multi-step serverless data workflow unmet.

Question 307

New documents use a different vector model from archived documents, and search quality drops despite identical index settings. What migration principle matters?

  1. Vector similarity metric
  2. Embedding model version consistency
  3. Text embedding creation
  4. Bedrock Knowledge Bases ingestion

Explanation

Correct Answer

B:

Mixing embeddings from incompatible model versions can make distance comparisons unreliable; in this scenario, re-embed indexed content or separate incompatible vector spaces before scoring new queries against the corpus.

Incorrect Answers

A:

Cosine similarity and other distance functions rank embeddings by geometric closeness under chosen semantics; its benefit is selecting a compatible similarity function for embeddings; that is different from preserving comparability after an embedding-model upgrade.

C:

Embedding models encode text into numeric vectors that represent semantic features for comparison; this can support converting product descriptions to vectors for similarity retrieval, but would not address the need for preserving comparability after an embedding-model upgrade.

D:

Bedrock Knowledge Bases can ingest supported sources, create embeddings and support retrieval workflows; this helps with preparing a managed document corpus for retrieval-augmented generation, not preserving comparability after an embedding-model upgrade.

Question 308

A schema conversion succeeds, but finance must ensure that customer balances and row counts agree between source and target. What final migration practice fits?

  1. AWS DMS change data capture
  2. AWS DMS Schema Conversion
  3. Dual-write migration validation
  4. Schema assessment report

Explanation

Correct Answer

C:

Validating row counts, checksums and business invariants detects differences between migrated and original data; in this scenario, compare representative data and key invariants at consistent snapshots before cutover, investigating any mismatch.

Incorrect Answers

A:

DMS CDC transfers ongoing supported source changes after baseline replication to keep a target synchronized; this helps with minimizing database migration cutover downtime, not proving a migrated target is consistent before traffic moves.

B:

AWS DMS Schema Conversion assesses and converts supported database schema objects between engines; although useful for migrating proprietary database objects to a different engine, it leaves the need for proving a migrated target is consistent before traffic moves unmet.

D:

A schema assessment highlights compatibility gaps and constructs requiring manual remediation during migration; its benefit is identifying unsupported conversion features early; that is different from proving a migrated target is consistent before traffic moves.

Question 309

An ingestion pipeline has an EMR cluster ready and needs to run a packaged Spark application after new data lands. What orchestration action fits?

  1. Redshift Data API execution
  2. Step Functions Glue run-job integration
  3. Athena StartQueryExecution
  4. Glue job start API
  5. EMR step submission

Explanation

Correct Answer

E:

EMR steps run specified data-processing tasks such as Spark jobs on a compatible cluster; in this scenario, submit an EMR step with the application arguments and monitor step state before marking the batch complete.

Incorrect Answers

A:

The Redshift Data API can submit SQL statements to compatible Redshift environments using authorized calls; its benefit is invoking warehouse SQL without maintaining a persistent client connection; that is different from launching Spark processing on an existing EMR cluster.

B:

A Step Functions Glue service integration can start a job and wait for its run to finish; the current need is for launching Spark processing on an existing EMR cluster, not coordinating downstream steps after a Glue job completes.

C:

Athena APIs can start SQL queries and return query execution IDs for polling results; this can support running a governed S3 SQL check from code, but would not address the need for launching Spark processing on an existing EMR cluster.

D:

The Glue StartJobRun API launches a configured ETL job and returns a run identifier for tracking; this helps with starting Glue ETL from an application workflow, not launching Spark processing on an existing EMR cluster.

Question 310

A data lake landing prefix includes `_temporary` directories that should never influence catalog schemas. Which crawler setting helps?

  1. Custom Glue classifier
  2. Glue crawler schema inference
  3. Crawler schema change policy
  4. Crawler exclusion pattern

Explanation

Correct Answer

D:

Crawler exclusions prevent matching paths and files from being considered during discovery; in this scenario, exclude staging and rejected paths so only intended data files contribute to the inferred dataset metadata.

Incorrect Answers

A:

A custom classifier influences how Glue crawlers recognize certain input formats and extract schemas; the current need is for excluding temporary and rejected files from inferred tables, not correctly recognizing a nonstandard delimited text format.

B:

A Glue crawler inspects supported source data and infers table structures for the Data Catalog; this can support discovering schemas from newly landed files, but would not address the need for excluding temporary and rejected files from inferred tables.

C:

Crawler change policies control how discovered schema differences update catalog objects; this helps with preventing surprise schema rewrites from new vendor files, not excluding temporary and rejected files from inferred tables.

AWS Certified Data Engineer - Associate DEA-C01 Practice Tests