Question 280
A clickstream alternates between quiet periods and traffic spikes; manual shard allocation has become error-prone. Which stream mode is suitable?
- RDS Proxy connection pooling
- Exponential backoff with jitter
- DynamoDB adaptive capacity
- Kinesis on-demand capacity
- Client-side concurrency limit
Explanation
Correct Answer
D:Kinesis on-demand mode adjusts stream capacity without manually planning shard counts for typical demand patterns; in this scenario, on-demand mode automates capacity management within its scaling behavior; monitor observed throughput and applicable limits.
Incorrect Answers
A:RDS Proxy pools application database connections and reduces connection churn against supported databases; this helps with protecting a relational database from short-lived connection bursts, not absorbing variable streaming ingestion demand without shard planning.
B:Backoff and jitter distribute API retries in time after throttling responses or transient errors; although useful for preventing synchronized retries after a shared API quota is exhausted, it leaves the need for absorbing variable streaming ingestion demand without shard planning unmet.
C:DynamoDB adaptive capacity can redistribute available throughput to heavily accessed partitions within service limits; it would target mitigating uneven partition access within a DynamoDB table, while the question is about absorbing variable streaming ingestion demand without shard planning.
E:A semaphore or worker cap bounds simultaneous API requests sent by an ingestion job; the current need is for absorbing variable streaming ingestion demand without shard planning, not keeping API requests below a known in-flight limit.
Question 281
Several teams use one S3 bucket but need distinct managed entry points with scoped policies. What feature helps?
- IAM RBAC role assignment
- Session tag authorization
- Lake Formation LF-Tag based access
- Cross-account bucket policy
- S3 access point policy
- IAM ABAC with principal tags
Explanation
Correct Answer
E:S3 access points offer distinct access controls and network policies for shared S3 datasets; in this scenario, create application-specific access points and policies, checking how they interact with the bucket and IAM permissions.
Incorrect Answers
A:Role-based access control groups permissions according to job functions or workload responsibilities; this can support giving production operators a stable predefined permission set, but would not address the need for exposing one bucket to applications with different data-access boundaries.
B:Session tags can carry approved attributes in assumed-role sessions for ABAC decisions; this helps with propagating per-session team attributes to authorization rules, not exposing one bucket to applications with different data-access boundaries.
C:LF-Tag authorization grants governed table access based on metadata classifications and expressions; although useful for authorizing lake assets by governed data classification, it leaves the need for exposing one bucket to applications with different data-access boundaries unmet.
D:An S3 bucket resource policy can allow a designated external account or role to access selected objects; it would target authorizing direct read access from another AWS account, while the question is about exposing one bucket to applications with different data-access boundaries.
F:Attribute-based access control compares principal and resource attributes through supported IAM condition keys; the current need is for exposing one bucket to applications with different data-access boundaries, not mapping ETL roles to same-team tagged resources.
Question 282
An ETL produces one file per Spark task and hundreds of tasks write only kilobytes each. What execution change helps?
- Columnar encoding dictionary
- Sorting output by query predicates
- S3 object compaction
- Avoiding overpartitioned Spark output
Explanation
Correct Answer
D:Excessive Spark partitions can create scheduling overhead and tiny output objects on low-volume workloads; in this scenario, reduce output partition count or compact afterward while monitoring worker parallelism and memory constraints.
Incorrect Answers
A:Dictionary encoding can efficiently represent repeated low-cardinality values in columnar storage formats; although useful for compressing repeated categorical attributes efficiently, it leaves the need for tuning parallel output granularity to data volume unmet.
B:Sorting records on frequently filtered fields can improve locality and block skipping for some engines; it would target increasing pruning efficiency inside large files, while the question is about tuning parallel output granularity to data volume.
C:Compaction combines many small analytic files into fewer appropriately sized objects; its benefit is reducing excessive query task overhead; that is different from tuning parallel output granularity to data volume.
Question 283
Glue Spark logs show executors repeatedly terminating on large group-by partitions. Which tuning direction is relevant?
- Executor memory tuning
- Glue error log inspection
- Data type casting validation
- Spark shuffle skew investigation
Explanation
Correct Answer
A:Spark executor memory and partition size affect whether large tasks spill or fail from memory pressure; in this scenario, inspect partition sizes, reduce per-task data pressure and size workers appropriately; extra driver memory alone will not repair executor OOM.
Incorrect Answers
B:Glue job failures produce execution logs that distinguish syntax, credentials, schema and runtime failures; although useful for locating the precise failing operation in Glue, it leaves the need for addressing executor out-of-memory errors on large partitions unmet.
C:Explicit type validation and casting prevent malformed input strings from corrupting typed transformations; it would target handling unexpected nonnumeric values in a numeric column, while the question is about addressing executor out-of-memory errors on large partitions.
D:Spark stage metrics can reveal straggler tasks and imbalanced shuffle partitions in a join; its benefit is finding a slow join caused by one hot key; that is different from addressing executor out-of-memory errors on large partitions.
Question 284
Every arriving 10 KB JSON object requires a quick field check and a generated metadata label. Which processing service fits?
- SageMaker Unified Studio data exploration
- Glue Spark preprocessing job
- Glue DataBrew recipe
- DataBrew profiling job
- Lambda event-driven preprocessing
Explanation
Correct Answer
E:Lambda can perform small validation or normalization tasks when an object-arrival event is received; in this scenario, invoke a bounded Lambda handler from the event route and send invalid records to a deliberate exception path.
Incorrect Answers
A:Unified Studio provides a governed collaborative environment for exploring and preparing datasets; its benefit is preparing datasets in a shared analytics project; that is different from normalizing small files immediately after landing.
B:Glue Spark jobs transform larger datasets with joins, repartitioning and schema mapping; the current need is for normalizing small files immediately after landing, not preparing large multi-file datasets for analytics.
C:DataBrew recipes apply documented preparation steps such as normalizing and cleaning tabular fields; this can support letting analysts prepare messy CSV columns without writing Spark code, but would not address the need for normalizing small files immediately after landing.
D:A DataBrew profile job measures dataset properties including value patterns, completeness and distributions; this helps with inspecting quality problems before deciding cleaning rules, not normalizing small files immediately after landing.
Question 285
Data owners need to find every downstream table affected by changing one source schema; relationships form a directed graph. Which algorithmic approach fits?
- Lambda /tmp ephemeral storage
- Lambda Amazon EFS mount
- Distributed partitioned compute
- Graph traversal algorithm
Explanation
Correct Answer
D:Graph traversal explores connected vertices and edges to follow relationships or dependencies; in this scenario, traverse downstream edges from the changed source to enumerate dependent assets, guarding against cycles and repeated visits.
Incorrect Answers
A:Lambda provides configurable temporary /tmp storage local to an execution environment; the current need is for following lineage dependencies between datasets, not buffering intermediate files during one function execution.
B:Lambda can mount an EFS file system through access points from appropriately networked functions; this can support sharing persistent files among multiple Lambda executions, but would not address the need for following lineage dependencies between datasets.
C:Distributed processing partitions data so workers handle portions in parallel and combine results as needed; this helps with parallelizing a large aggregation beyond one machine, not following lineage dependencies between datasets.
Question 286
A feature store is updated in DynamoDB, and an enrichment worker must react only when table items change. Which stream source is appropriate?
- Amazon MSK
- DynamoDB Streams
- Amazon Kinesis Data Streams
- Amazon Data Firehose
Explanation
Correct Answer
B:DynamoDB Streams exposes item-level changes made to a DynamoDB table for downstream processors; in this scenario, enable the table stream with the needed view type and consume change records; it is not a general event-ingest endpoint.
Incorrect Answers
A:Amazon MSK runs managed Apache Kafka brokers while retaining Kafka client and topic semantics; the current need is for reacting to inserts and updates of table items, not migrating existing Kafka producers without rewriting clients.
C:A data stream retains ordered records in shards for applications that read at their own pace; this helps with keeping a replayable clickstream for independent consumers, not reacting to inserts and updates of table items.
D:Firehose buffers streaming records and delivers them to configured analytics destinations; although useful for continuously depositing events into an S3 data lake, it leaves the need for reacting to inserts and updates of table items unmet.
Question 287
Device feeds arrive on schedule but values frequently conflict with approved calibration ranges. Which data-quality dimension needs focus?
- Data veracity validation
- High-volume batch architecture
- High-velocity streaming architecture
- Schema-flexible variety handling
Explanation
Correct Answer
A:Veracity concerns trustworthiness and reliability of incoming information rather than its quantity or speed; in this scenario, validate accuracy, source credibility and anomaly patterns, because the main risk is whether the recorded data can be trusted.
Incorrect Answers
B:Volume describes the total quantity of data that must be stored and processed; although useful for scaling storage and parallelism for terabytes of historical data, it leaves the need for checking suspicious and conflicting source records unmet.
C:Velocity describes how rapidly data arrives and how quickly it needs to be handled; it would target coping with millions of events arriving per second, while the question is about checking suspicious and conflicting source records.
D:Variety covers multiple structures and types including relational rows, JSON, images and free text; its benefit is ingesting incompatible structured and unstructured formats; that is different from checking suspicious and conflicting source records.
Question 288
Operations require fast text search for an unusual exception token across many normalized pipeline logs. Which service helps?
- CloudWatch Logs subscription forwarding
- OpenSearch Service indexed audit exploration
- CloudWatch Logs Insights query
- Athena over S3 CloudTrail logs
- CloudTrail Lake SQL query
Explanation
Correct Answer
B:OpenSearch can index log documents to support interactive search and visualization; in this scenario, index chosen log fields and query them with the search interface, managing lifecycle and sensitive log access.
Incorrect Answers
A:Subscription filters can deliver matching CloudWatch log events to downstream processing destinations; this can support streaming application audit records to a central pipeline, but would not address the need for searching repeated substrings across a large normalized log index.
C:Logs Insights performs interactive filtered searches and aggregations over CloudWatch application log groups; although useful for counting application exception types in Lambda logs, it leaves the need for searching repeated substrings across a large normalized log index unmet.
D:Athena can query CloudTrail files in S3 after defining compatible schemas and partitions; it would target joining archived API logs with lake reference datasets, while the question is about searching repeated substrings across a large normalized log index.
E:CloudTrail Lake supports SQL-based analysis over events collected in event data stores; its benefit is investigating account activity with structured SQL; that is different from searching repeated substrings across a large normalized log index.
Question 289
A company needs standard API audit collection from a fleet of analytics accounts without manually configuring each member. What setup fits?
- CloudTrail management-event records
- CloudTrail log file integrity validation
- CloudTrail S3 object data events
- CloudTrail organization trail
Explanation
Correct Answer
D:An organization trail can collect supported management activity centrally for accounts in an AWS Organization; in this scenario, create and secure an organization-wide trail and centrally manage its logging destinations and retention controls.
Incorrect Answers
A:Management events track AWS control-plane actions such as PutBucketPolicy and CreateRole; this can support finding who changed resource permissions, but would not address the need for covering many accounts with one audit policy.
B:Trail log file integrity validation provides digest files that can help detect tampering of delivered log files; this helps with proving stored audit logs were not silently altered, not covering many accounts with one audit policy.
C:CloudTrail can capture selected S3 object-level APIs as data events when explicitly configured; although useful for identifying who downloaded restricted data-lake objects, it leaves the need for covering many accounts with one audit policy unmet.
Question 290
A scheduled job sometimes fails to start because its target invocation is temporarily unavailable. What scheduler behavior should be configured?
- Airflow catchup configuration
- Step Functions task token callback
- Glue workflow run properties
- EventBridge Scheduler retry policy
Explanation
Correct Answer
D:Scheduler targets can have retry policies and dead-letter queue configuration for failed delivery attempts; in this scenario, set bounded retries and a dead-letter destination while ensuring the invocation role permits the target operation.
Incorrect Answers
A:Airflow catchup behavior determines whether a DAG schedules missed historical intervals when enabled; this can support preventing a flood of old scheduled DAG runs, but would not address the need for recovering from temporary failure to invoke a scheduled target.
B:A task token lets a state-machine task wait for an external system to signal completion; this helps with pausing a workflow for asynchronous external approval, not recovering from temporary failure to invoke a scheduled target.
C:Glue workflow run properties allow pipeline components to share run-scoped configuration values; although useful for passing an ingestion batch identifier between Glue jobs, it leaves the need for recovering from temporary failure to invoke a scheduled target unmet.