Question 150
An application deletes a key in a versioned S3 bucket, but earlier data versions remain stored. Which recovery operation fits?
- DynamoDB point-in-time recovery
- DynamoDB TTL attribute
- S3 bucket versioning
- S3 delete-marker recovery
- S3 MFA Delete
Explanation
Correct Answer
D:Deleting the current delete marker can reveal an older object version in a versioned bucket; in this scenario, locate and remove the current delete marker where appropriate, exposing the preserved prior object as current again.
Incorrect Answers
A:PITR supports continuous backups of a DynamoDB table within its supported recovery window; this helps with restoring a table after accidental bulk writes, not restoring a logically deleted current object.
B:DynamoDB TTL uses a specified numeric epoch-seconds attribute to expire items asynchronously; although useful for removing stale session records from DynamoDB, it leaves the need for restoring a logically deleted current object unmet.
C:Versioning preserves separate revisions of an object key to aid recovery from overwrites and deletes; it would target recovering a previously overwritten object revision, while the question is about restoring a logically deleted current object.
E:MFA Delete adds an MFA requirement to certain versioning-related delete operations when configured by the root account; the current need is for restoring a logically deleted current object, not adding strong protection to permanent deletion of object versions.
Question 151
A password change currently requires updating every ETL script that reads the same relational source. What credential arrangement helps?
- SageMaker Catalog business asset
- SageMaker Unified Studio project
- Glue JDBC connection definition
- Glue secret reference
- SageMaker domain unit
Explanation
Correct Answer
D:A Glue connection can use a secret for database credentials when configured and permitted; in this scenario, reference credentials through a managed secret with controlled rotation rather than embedding passwords in job arguments.
Incorrect Answers
A:SageMaker Catalog can provide business-oriented discovery and descriptions of governed data assets; its benefit is helping analysts discover an approved business dataset; that is different from rotating shared source credentials without editing multiple jobs.
B:A Unified Studio project provides a collaborative workspace and governed resources for data and analytics activities; the current need is for rotating shared source credentials without editing multiple jobs, not organizing a data team's approved work context.
C:A Glue connection saves reusable endpoint, authentication and networking information for supported sources; this can support reusing database connectivity settings across Glue jobs, but would not address the need for rotating shared source credentials without editing multiple jobs.
E:Domain units can organize teams and governance boundaries within a SageMaker Unified Studio domain; although useful for delegating catalog administration to a business unit, it leaves the need for rotating shared source credentials without editing multiple jobs unmet.
Question 152
A team operates dozens of Python Airflow DAGs with custom operators and wants AWS to manage the Airflow environment. Which service fits?
- EventBridge rule
- Amazon MWAA DAG
- AWS Step Functions state machine
- Glue workflow
- Lambda direct invocation
Explanation
Correct Answer
B:Amazon MWAA provides managed Apache Airflow for dependency graphs expressed as DAGs; in this scenario, mWAA preserves DAG and operator concepts while AWS manages the surrounding Airflow infrastructure and control plane.
Incorrect Answers
A:EventBridge matches emitted events and routes them to targets without tracking a multi-step workflow state; this helps with routing a completed-upload event to the next service, not running an established Airflow-based cross-platform DAG.
C:Step Functions coordinates service tasks with explicit states, branching, errors and retries; it would target orchestrating a serverless ETL with conditional paths, while the question is about running an established Airflow-based cross-platform DAG.
D:A Glue workflow links Glue triggers, crawlers and jobs into a managed ETL dependency sequence; its benefit is sequencing a Glue-only crawl and transform chain; that is different from running an established Airflow-based cross-platform DAG.
E:One Lambda function can synchronously or asynchronously invoke another for small, bounded tasks; the current need is for running an established Airflow-based cross-platform DAG, not initiating a single short helper action from code.
Question 153
Glue workers are in a VPC but receive timeouts on the database TCP port. What network control should be checked first?
- S3 gateway endpoint
- Security group inbound rule
- VPC interface endpoint
- NAT gateway stable egress IP
Explanation
Correct Answer
B:A security group rule allows specified inbound protocols, ports and sources to a resource interface; in this scenario, permit the job's security group as a source on the database listener port while retaining narrow ingress access.
Incorrect Answers
A:An S3 gateway endpoint adds a VPC route to S3 through the AWS private network path; the current need is for allowing a Glue job to reach a private JDBC database, not reading S3 objects from a private subnet without a NAT gateway.
C:An interface VPC endpoint offers private connectivity to supported AWS service APIs using endpoint ENIs; this helps with avoiding public internet transit for a supported AWS API, not allowing a Glue job to reach a private JDBC database.
D:A NAT gateway in a public subnet can provide predictable public egress through its Elastic IP; although useful for providing fixed outbound addresses for a private-subnet API poller, it leaves the need for allowing a Glue job to reach a private JDBC database unmet.
Question 154
A Glue ETL spends most time in a large join and some tasks spill heavily to disk. Which diagnostic data is most useful?
- Glue Spark stage metrics
- Athena scanned-data metrics
- Kinesis shard throughput metrics
- Redshift EXPLAIN plan
Explanation
Correct Answer
A:Spark task and stage metrics reveal shuffle size, skew, spill and slow executors; in this scenario, review Spark stage and executor metrics to determine whether skew or memory pressure causes the slow stage.
Incorrect Answers
B:Athena query execution statistics report the volume of data scanned and runtime behavior; its benefit is explaining a sudden rise in query cost; that is different from identifying why one Spark stage dominates the run.
C:Stream shard metrics reveal write or read throughput saturation and throttling within shard limits; the current need is for identifying why one Spark stage dominates the run, not diagnosing write throttling on a hot partition key.
D:EXPLAIN shows a planned Redshift query strategy, including joins, scans and potential data movement; this can support identifying costly redistribution in warehouse SQL, but would not address the need for identifying why one Spark stage dominates the run.
Question 155
A weather vendor offers only a REST endpoint and requires a fetch every ten minutes. What intake design is appropriate?
- API Gateway with Lambda
- Checkpointed API pagination
- Lambda scheduled API poller
- Secrets Manager for API tokens
Explanation
Correct Answer
C:Lambda code can invoke third-party REST APIs on a schedule when network access is available; in this scenario, invoke a small API client on schedule and persist fetched records, handling tokens, pagination and vendor quotas.
Incorrect Answers
A:API Gateway accepts HTTPS requests and can invoke Lambda for backend request processing; it would target exposing a managed HTTPS entry point for ingested events, while the question is about polling a vendor API at bounded intervals.
B:A durable cursor or page token enables an ingestion process to resume an API listing incrementally; its benefit is resuming paginated ingestion after partial failures; that is different from polling a vendor API at bounded intervals.
D:Secrets Manager stores API credentials and can support rotation workflows for supported secrets; this can support managing third-party API credentials without embedding them in code, but would not address the need for polling a vendor API at bounded intervals.
Question 156
A risk engine needs Redis data structures and fast persistent key-value operations without treating the store as an expendable cache. Which service fits?
- Amazon MemoryDB for Redis
- Amazon ElastiCache cache
- Aurora PostgreSQL pgvector HNSW index
- Amazon OpenSearch Service search index
Explanation
Correct Answer
A:MemoryDB offers a durable Redis-compatible in-memory database for low-latency key-value access; in this scenario, memoryDB provides the Redis-compatible interface with durable storage semantics suited to the application's system-of-record need.
Incorrect Answers
B:ElastiCache provides managed in-memory caching to accelerate access to data stored durably elsewhere; this can support caching read-heavy reference values without changing the source of truth, but would not address the need for combining Redis-compatible operations with a durable primary database.
C:Aurora PostgreSQL with supported pgvector functionality can index embeddings for approximate similarity search; this helps with finding similar embeddings within relational application data, not combining Redis-compatible operations with a durable primary database.
D:OpenSearch Service supports full-text search and search-oriented indexing of documents; although useful for enabling relevance-ranked keyword search over ingested articles, it leaves the need for combining Redis-compatible operations with a durable primary database unmet.
Question 157
The analytics team needs a recurring transfer of Salesforce account objects into Amazon S3 without building an API client. What best handles ingestion?
- AWS Glue JDBC connection
- Amazon AppFlow
- S3 object batch processing with Glue
- AWS DMS full load
Explanation
Correct Answer
B:AppFlow transfers data between supported SaaS applications and AWS destinations with managed connectors; in this scenario, a configured flow handles the SaaS connector, field mapping and scheduled transfer to the chosen destination.
Incorrect Answers
A:Glue can read JDBC-backed sources through a connection with network and credential settings; it would target extracting scheduled rows from a private relational database, while the question is about importing CRM objects from a supported SaaS application.
C:Glue batch ETL processes stored S3 objects when a scheduled job runs against their paths; the current need is for importing CRM objects from a supported SaaS application, not reprocessing a monthly collection of stored flat files.
D:AWS DMS can copy existing database tables as an initial migration or replication phase; this can support taking a one-time baseline of source tables, but would not address the need for importing CRM objects from a supported SaaS application.
Question 158
Mobile telemetry adds optional nested keys frequently and analysts need to retain fields before normalizing them. What landing format fits?
- Apache Parquet output
- JSON preservation for variable schema
- Compression-aware partitioned Parquet
- Apache Avro for row-oriented exchange
Explanation
Correct Answer
B:JSON can preserve nested and changing attributes that are awkward in a rigid flat-row schema; in this scenario, preserve raw JSON events first, then curate schema-specific tables when contracts and data usage become clearer.
Incorrect Answers
A:Parquet stores column-oriented data with encoding and compression suitable for selective analytics; this helps with reducing scan bytes for repeated analytic queries, not preserving semi-structured event attributes during initial landing.
C:Compression and sensible file sizing improve Parquet read efficiency without creating countless tiny files; it would target avoiding a data lake filled with tiny output objects, while the question is about preserving semi-structured event attributes during initial landing.
D:Avro serializes row-oriented records with schema-based encoding suitable for data interchange; its benefit is exchanging schema-described records across heterogeneous consumers; that is different from preserving semi-structured event attributes during initial landing.
Question 159
A Glue job's subnet routing is correct, but strict worker security-group egress blocks its JDBC port. Which rule requires review?
- Private VPC endpoint policy
- Database security-group ingress
- Network ACL rule order
- Glue worker security-group egress
Explanation
Correct Answer
D:Security-group egress rules govern which destinations Glue worker ENIs may contact; in this scenario, permit the specified destination and port on the job-side security group while retaining restrictive outbound access elsewhere.
Incorrect Answers
A:VPC endpoint policies can restrict supported API actions and resource access through a particular endpoint; it would target limiting S3 API access through a specific gateway endpoint, while the question is about allowing a job to initiate approved outbound database connections.
B:A VPC security group can restrict database listener access by source security group and destination port; its benefit is permitting only Glue ETL workers to reach RDS over TCP; that is different from allowing a job to initiate approved outbound database connections.
C:Network ACLs are stateless subnet controls with numbered allow and deny rules evaluated in order; the current need is for allowing a job to initiate approved outbound database connections, not debugging return traffic dropped at a subnet boundary.
Question 160
A Redshift ETL appears idle while other sessions run, and logs suggest a lock wait on a target table. What should be inspected?
- Partitioned load scheduling
- Redshift lock monitoring
- Shorter database transaction
- Relational isolation-level selection
Explanation
Correct Answer
B:Redshift system views and query monitoring expose sessions waiting on resource or transaction locks; in this scenario, examine active queries, transactions and lock status to identify the blocker before terminating or rescheduling workloads.
Incorrect Answers
A:Loading separate table partitions at coordinated intervals can reduce contention with user-facing operations; this can support avoiding bulk-load conflicts with live warehouse queries, but would not address the need for diagnosing warehouse query waits caused by locks.
C:Reducing transaction length releases row and table locks sooner in conventional relational engines; although useful for limiting lock contention from long-running batch updates, it leaves the need for diagnosing warehouse query waits caused by locks unmet.
D:Transaction isolation controls the concurrency anomalies a database transaction permits under its engine semantics; it would target choosing consistency guarantees for simultaneous source updates, while the question is about diagnosing warehouse query waits caused by locks.