Design High-Performing Architectures

100. A read-heavy application on Amazon RDS shows high CPU caused by repeated identical queries. Which solution most directly reduces load on the database?

1. A Convert the database to a Multi-AZ deployment and direct read traffic to the standby.
2. B Move the application to a larger Amazon EC2 instance type.
3. C Deploy an Amazon ElastiCache cluster and cache the results of the repeated queries.
4. D Increase the allocated storage and change the volume to Provisioned IOPS.

101. Amazon Athena queries against millions of small JSON files in Amazon S3 are slow and expensive. Which combination of steps provides the largest improvement? (Select TWO.)

1. A Convert the data to Apache Parquet and partition it by the columns used in query filters.
2. B Change the S3 storage class of the files to S3 Standard-Infrequent Access.
3. C Increase the Athena query timeout so long queries complete.
4. D Compact the small files into objects of roughly 128 MB or larger.
5. E Enable S3 Transfer Acceleration on the bucket.

102. An Amazon EBS volume backing a database shows high queue depth and latency while instance CPU remains low. Which solution meets these requirements?

1. A Change the volume to io2 or gp3 and provision additional IOPS to match the workload.
2. B Attach a second volume of the same type and stripe the data across both.
3. C Change the instance to a type with more vCPUs and more memory.
4. D Enable encryption at rest on the existing volume.

103. An API receives repeated identical GET requests for reference data that changes twice a day. Which solution most reduces load on the backend?

1. A Enable Amazon API Gateway caching on the stage with a time to live shorter than the update interval.
2. B Increase the memory allocated to the AWS Lambda function behind the API.
3. C Deploy the backend across an additional Availability Zone.
4. D Convert the REST API to a WebSocket API so clients receive pushed updates.

104. An application performs heavy read traffic against Amazon RDS, and reads tolerate data that is a few seconds stale. Which solution most directly increases read throughput?

1. A Convert the instance to a Multi-AZ deployment and direct read traffic to the standby.
2. B Increase the automated backup retention period and enable enhanced monitoring.
3. C Change the instance storage to Provisioned IOPS and increase the allocated capacity.
4. D Create one or more read replicas and direct read traffic to their endpoints.

105. Users on other continents experience slow uploads of large files to an Amazon S3 bucket. Which solution meets these requirements?

1. A Enable S3 Transfer Acceleration on the bucket and have clients upload through the accelerated endpoint.
2. B Enable S3 Intelligent-Tiering on the bucket to optimize the storage class.
3. C Enable S3 Versioning on the bucket so failed uploads can be retried.
4. D Configure S3 Cross-Region Replication to a bucket nearer the users.

106. A database requires sustained 60,000 IOPS with consistent sub-millisecond latency from its Amazon EBS storage. Which solution meets these requirements?

1. A Attach an io2 Block Express volume with the required IOPS provisioned.
2. B Attach a gp2 volume sized so that its baseline IOPS reaches the requirement.
3. C Attach an st1 throughput-optimized HDD volume sized for the workload.
4. D Attach several sc1 cold HDD volumes and stripe the data across them.

107. A Linux workload built on open-source runtimes must reduce compute cost while maintaining performance. Which solution is most likely to meet these requirements?

1. A Change the attached EBS volumes from gp2 to gp3.
2. B Enable detailed CloudWatch monitoring on the existing instances.
3. C Migrate the workload to Graviton-based instance types and validate performance after the change.
4. D Migrate the workload to a larger instance size within the same instance family.

108. Users worldwide report slow downloads of large static assets served from Amazon EC2 instances in one Region. Which solution provides the largest latency improvement?

1. A Create an Amazon CloudFront distribution with the EC2 instances as the origin.
2. B Enable enhanced networking on the existing EC2 instances.
3. C Move the EC2 instances to a different subnet in the same Region.
4. D Launch additional EC2 instances of a larger instance type in the same Region.

109. A CPU-bound video transcoding job runs for several hours each night and can be interrupted and resumed. Which solution provides the best price for performance?

1. A Run the job on general purpose instances covered by three-year Standard Reserved Instances.
2. B Run the job on AWS Lambda functions invoked in parallel.
3. C Run the job on memory-optimized On-Demand Instances sized for the workload.
4. D Run the job on compute-optimized Spot Instances in an Auto Scaling group with a mixed instances policy.

110. A tightly coupled high performance computing workload requires very low latency and high packet-per-second performance between nodes. Which combination of steps meets these requirements? (Select TWO.)

1. A Distribute the instances evenly across three Availability Zones.
2. B Launch the instances into a cluster placement group.
3. C Launch the instances as Spot Instances across several instance families.
4. D Route inter-node traffic through a NAT gateway in a public subnet.
5. E Attach an Elastic Fabric Adapter to each instance on a supported instance type.

111. A containerized workload has variable demand, and the team wants per-task isolation without provisioning or patching instances. Which solution meets these requirements?

1. A Run the tasks on Amazon ECS with the AWS Fargate launch type.
2. B Run the tasks on Amazon ECS with the EC2 launch type and a Spot capacity provider.
3. C Run the tasks on Amazon EC2 instances in a cluster placement group with Docker installed.
4. D Run the tasks on Amazon ECS with the EC2 launch type and a managed node group.

112. An AWS Lambda function performing CPU-intensive image processing runs slower than required, and its memory usage is close to the configured allocation. Which solution meets these requirements?

1. A Increase the function's timeout so the invocation has more time to complete.
2. B Enable provisioned concurrency so execution environments are initialized in advance.
3. C Configure the function to run inside a VPC with an interface endpoint.
4. D Increase the function's memory allocation so that proportionally more CPU is allocated.

113. An Amazon DynamoDB table is throttling because most requests target a small number of partition key values. Which combination of steps meets these requirements? (Select TWO.)

1. A Create a local secondary index on the frequently queried attribute.
2. B Deploy DynamoDB Accelerator in front of the table to absorb repeated reads of the hot items.
3. C Convert the table to a global table with a replica in a second Region.
4. D Increase the table's provisioned write capacity until throttling stops.
5. E Redesign the partition key to a higher-cardinality value, adding a write-sharding suffix where necessary.

114. A gaming leaderboard must return the top 100 players with sub-millisecond latency at very high request rates. Which solution meets these requirements?

1. A Store the scores in Amazon RDS and read the top range from a read replica.
2. B Store the scores in Amazon S3 and query the top range with Amazon Athena.
3. C Store the scores in an Amazon ElastiCache for Redis sorted set and read the top range from it.
4. D Store the scores as a JSON object in Amazon S3 and parse it on each request.

115. A relational workload has long idle periods punctuated by unpredictable bursts, and the team wants to avoid paying for idle capacity. Which solution meets these requirements?

1. A Deploy an Aurora Serverless v2 cluster and configure a capacity range that covers the bursts.
2. B Deploy a provisioned Aurora cluster sized for the largest expected burst.
3. C Deploy an Amazon DynamoDB table configured with on-demand capacity mode.
4. D Deploy an Amazon RDS instance on a burstable instance class sized for the average load.

116. A read-heavy Amazon DynamoDB workload repeatedly requests the same small set of items and must return them in under one millisecond. Which solution meets these requirements?

1. A Deploy a DynamoDB Accelerator cluster and route the read requests through it.
2. B Create a global secondary index on the table's partition key and query the index.
3. C Increase the table's provisioned read capacity by an order of magnitude.
4. D Enable DynamoDB Streams and maintain a copy of the hot items in Amazon S3.

117. An analytics team runs complex joins over 40 TB of structured data with high concurrency and requires consistent sub-minute query times. Which solution meets these requirements?

1. A Load the data into Amazon DynamoDB and create global secondary indexes for each query.
2. B Load the data into Amazon Redshift and define distribution and sort keys that match the join and filter columns.
3. C Load the data into Amazon RDS for PostgreSQL on a memory-optimized instance class.
4. D Query the data in place with Amazon Athena and partition the files by the filter columns.

118. An API backed by AWS Lambda experiences unacceptable cold-start latency during a predictable morning traffic spike. Which solution meets these requirements?

1. A Increase the function timeout so slow invocations have time to complete.
2. B Configure provisioned concurrency on the function and schedule it to increase before the spike.
3. C Configure reserved concurrency on the function to guarantee it capacity.
4. D Place an Amazon SQS queue in front of the function to buffer requests.

119. A dynamic, non-cacheable API served from one Region must have the lowest possible latency for global users and must fail over to a second Region within seconds. Which solution meets these requirements?

1. A Create an AWS Global Accelerator accelerator with endpoint groups in both Regions.
2. B Create an Amazon CloudFront distribution with caching disabled and one Regional origin.
3. C Enable cross-zone load balancing on the Application Load Balancer in the primary Region.
4. D Create Amazon Route 53 weighted records for both Regional endpoints.

120. An Amazon CloudFront distribution serves a large catalogue of infrequently requested objects, and load on the origin remains high. Which solution meets these requirements?

1. A Enable Origin Shield on the distribution so edge cache misses are consolidated through one caching layer.
2. B Increase the distribution's default time to live so objects remain cached for longer.
3. C Add a second origin to the distribution and balance requests between them.
4. D Enable field-level encryption on the distribution to reduce the payload size.

121. An application must connect an on-premises data center to a VPC with consistent bandwidth and predictable latency. Which solution meets these requirements?

1. A Configure an AWS Site-to-Site VPN between the data center and the VPC.
2. B Attach the VPC to an AWS Transit Gateway and route on-premises traffic through it.
3. C Configure AWS Global Accelerator with the VPC resources as endpoints.
4. D Provision an AWS Direct Connect connection between the data center and the VPC.

122. An architecture must connect 30 VPCs across two Regions with centralized routing and segmentation between environments. Which solution meets these requirements?

1. A Create VPC peering connections between every pair of VPCs and add routes in each route table.
2. B Publish each VPC's services through AWS PrivateLink endpoints in every other VPC.
3. C Deploy an AWS Transit Gateway in each Region, peer them, and use separate route tables per environment.
4. D Consolidate all workloads into a single VPC with separate subnets per environment.

123. A batch analytics job must process 50 TB of data in Amazon S3 with Apache Spark, and the team does not want to manage clusters. Which solution meets these requirements?

1. A Launch an Amazon EMR cluster before each run and terminate it when the job completes.
2. B Run Spark on a single large Amazon EC2 instance sized for the dataset.
3. C Submit the job to AWS Batch on Amazon EC2 with Spark installed in the container image.
4. D Submit the job to Amazon EMR Serverless and let the service provision capacity automatically.

124. A clickstream must be consumed by several independent teams, each reading the full record set from the beginning of the retention window at its own pace. Which solution meets these requirements?

1. A Publish the records to an AWS Step Functions workflow that invokes each team's processor.
2. B Publish the records to an Amazon Kinesis Data Stream and have each team run its own consumer.
3. C Publish the records to an Amazon SQS queue and have each team poll the queue.
4. D Publish the records to an Amazon SNS topic and subscribe each team's endpoint.

125. A company must transfer 200 TB from an on-premises NFS share to Amazon S3 over a 10 Gbps dedicated link while preserving file metadata. Which solution meets these requirements?

1. A Use the AWS CLI to copy the files to Amazon S3 in parallel batches.
2. B Ship the data on AWS Snowball Edge devices and import it into Amazon S3.
3. C Deploy AWS DataSync agents on premises and create a task that transfers the share to Amazon S3.
4. D Deploy AWS Storage Gateway in file gateway mode and copy the files through the local cache.

126. Streaming records must be delivered to Amazon S3 in near real time, converted to Apache Parquet, and partitioned by ingestion date, with no servers to manage. Which solution meets these requirements?

1. A Create an Amazon Kinesis Data Stream and write the records to Amazon S3 from a consumer application.
2. B Create an Amazon Data Firehose delivery stream with record format conversion and dynamic partitioning enabled.
3. C Create an Amazon EMR cluster running Spark Structured Streaming that writes to Amazon S3.
4. D Create an AWS Glue streaming ETL job that reads the records and writes them to Amazon S3.

127. An analytics team must query several terabytes of CSV files in Amazon S3 with SQL, and query cost must be minimized. Which solution meets these requirements?

1. A Query the CSV files directly with Amazon Athena and increase the workgroup's data scan limit.
2. B Load the CSV files into an Amazon RDS instance and query them with SQL.
3. C Convert the files to a partitioned columnar format such as Apache Parquet and query them with Amazon Athena.
4. D Compress the CSV files with gzip and query them with Amazon Athena.

128. A company must build a data lake in Amazon S3 with centrally managed table and column permissions for analysts using Amazon Athena and Amazon Redshift Spectrum. Which solution meets these requirements?

1. A Attach IAM policies to each analyst that allow s3:GetObject on the relevant object keys.
2. B Register the S3 locations with AWS Lake Formation and grant table and column permissions to the analyst principals.
3. C Create a separate S3 bucket for each analyst group and copy the relevant tables into it.
4. D Apply an S3 bucket policy for each analyst group that restricts access to the relevant prefixes.

129. An ETL job must run each night to transform data in Amazon S3 and write partitioned output, with no cluster to provision. Which solution meets these requirements?

1. A Invoke an AWS Lambda function on a schedule to transform the full dataset in one execution.
2. B Launch an Amazon EMR cluster each night, run the transformation, and terminate the cluster.
3. C Create an AWS Glue ETL job with a scheduled trigger that writes the partitioned output to Amazon S3.
4. D Run the transformation on an Amazon EC2 instance started by a scheduled action.

130. A workload requires the highest possible IOPS from a single Amazon EBS volume. Which volume type is appropriate?

1. A gp3, which provides a configurable baseline.
2. B st1, which is throughput optimized.
3. C sc1, which is cold HDD storage.
4. D io2 Block Express, which supports the highest provisioned IOPS.

131. A data lake's objects have unpredictable access patterns, and storage cost must be minimized without retrieval delays. Which storage class is appropriate?

1. A S3 Glacier Flexible Retrieval, which imposes retrieval delays.
2. B S3 One Zone-Infrequent Access, which stores in a single Availability Zone.
3. C S3 Intelligent-Tiering, which moves objects between access tiers automatically without retrieval charges for the main tiers.
4. D S3 Standard, which charges frequent-access rates for all objects.

132. A high performance computing workload requires a shared file system with sub-millisecond latency and very high throughput. Which solution meets these requirements?

1. A Amazon EFS in General Purpose mode.
2. B Amazon EBS gp3 volumes on each instance.
3. C Amazon S3 accessed through the API.
4. D Amazon FSx for Lustre, optionally linked to an S3 bucket.

133. A workload writes temporary scratch data that does not need to survive instance termination and requires the lowest latency. Which storage is appropriate?

1. A Amazon EBS gp3 volumes with a delete-on-termination flag.
2. B Amazon S3 with a lifecycle expiry.
3. C Instance store volumes, which are physically attached and lost on termination.
4. D Amazon EFS with a lifecycle policy.

134. A workload runs in short bursts triggered by events and must incur no cost when idle. Which compute option is appropriate?

1. A An ECS service on EC2 with a minimum of one task.
2. B An EC2 instance in an Auto Scaling group with a minimum of one.
3. C A dedicated EC2 instance sized for the burst.
4. D AWS Lambda, which bills per invocation and duration.

135. A containerized workload must run without the team managing any EC2 instances. Which solution meets these requirements?

1. A Run the containers on Amazon EKS with managed node groups.
2. B Run the containers on Amazon ECS with an EC2 capacity provider.
3. C Run the containers on EC2 instances with Docker installed.
4. D Run the containers on AWS Fargate.

136. A machine learning training workload can tolerate interruption and must minimize cost. Which purchasing option is appropriate?

1. A Reserved Instances, which require a term commitment.
2. B On-Demand Instances, which have no commitment.
3. C Dedicated Hosts, which provide hardware isolation.
4. D Spot Instances, which use spare capacity at a deep discount and may be reclaimed.

137. An application's compute must scale out before a known daily traffic peak rather than reacting to it. Which solution meets these requirements?

1. A Configure scheduled scaling to increase capacity ahead of the peak.
2. B Increase the Auto Scaling group's maximum size.
3. C Configure target tracking on CPU utilization.
4. D Configure a step scaling policy with a short cooldown.

138. A relational database's read traffic has grown and is affecting write performance. Which solution meets these requirements?

1. A Add read replicas and direct read queries to them.
2. B Enable Multi-AZ for the database.
3. C Take more frequent snapshots.
4. D Increase the database's provisioned IOPS.

139. A DynamoDB table's read traffic is dominated by repeated reads of the same items requiring microsecond latency. Which solution meets these requirements?

1. A Add DynamoDB Accelerator as an in-memory cache in front of the table.
2. B Enable DynamoDB Streams on the table.
3. C Increase the table's provisioned read capacity.
4. D Add a global secondary index on the accessed attribute.

140. An application requires a database supporting complex SQL joins with automatic scaling of storage up to 128 TB. Which solution meets these requirements?

1. A Amazon Neptune, which is a graph database.
2. B Amazon DynamoDB, which is key-value and document oriented.
3. C Amazon Aurora, which scales storage automatically and supports complex SQL.
4. D Amazon ElastiCache, which is an in-memory cache.

141. A DynamoDB table must support queries on an attribute that is not part of its primary key, with its own partition key. Which solution meets these requirements?

1. A Create a local secondary index with the required attribute.
2. B Create a global secondary index with the required partition key.
3. C Change the table's partition key to the attribute.
4. D Scan the table and filter on the attribute.

142. Static content must be delivered to global users with low latency. Which solution meets these requirements?

1. A Serve the content from a larger EC2 instance in one Region.
2. B Compress the content before serving it.
3. C Distribute the content through Amazon CloudFront so it is cached at edge locations near users.
4. D Replicate the content to every Region and route by geolocation.

143. A TCP-based application must be reachable through static IP addresses with traffic entering the AWS network near users. Which solution meets these requirements?

1. A An Application Load Balancer with a static DNS name.
2. B Amazon CloudFront, which is optimized for HTTP content delivery.
3. C Amazon Route 53 latency-based routing.
4. D AWS Global Accelerator, which provides static anycast addresses and routes over the AWS backbone.

144. Instances in a cluster must communicate with the lowest possible network latency between them. Which solution meets these requirements?

1. A Launch the instances in a cluster placement group within a single Availability Zone.
2. B Launch the instances in a spread placement group across Availability Zones.
3. C Launch the instances with Elastic IP addresses.
4. D Launch the instances in different Availability Zones for resilience.

145. An application must reach Amazon S3 from a private subnet without traversing the internet, at the lowest cost. Which solution meets these requirements?

1. A Route through a NAT gateway.
2. B Assign public IP addresses to the instances.
3. C Create a gateway VPC endpoint for Amazon S3 and add a route to the subnet's route table.
4. D Create an interface VPC endpoint for Amazon S3.

146. Streaming data must be delivered to Amazon S3 in near real time with no servers to manage. Which solution meets these requirements?

1. A Amazon EMR with Spark Structured Streaming.
2. B Amazon Data Firehose, which buffers and delivers streaming data to Amazon S3.
3. C Amazon SQS with a Lambda consumer writing to Amazon S3.
4. D Amazon Kinesis Data Streams with a custom consumer application.

147. A serverless solution must query data stored in Amazon S3 using standard SQL. Which solution meets these requirements?

1. A Amazon EMR, which requires a cluster.
2. B Amazon RDS, which requires the data to be loaded.
3. C Amazon Athena, which queries S3 data with SQL and no infrastructure.
4. D Amazon Redshift, which requires a provisioned cluster.

148. A serverless ETL job must discover the schema of data in Amazon S3 and transform it. Which solution meets these requirements?

1. A AWS Glue with a crawler to populate the Data Catalog and a job to perform the transformation.
2. B An EC2 instance running a transformation script.
3. C Amazon Athena with a CREATE TABLE statement.
4. D Amazon EMR with a Spark job.

149. A workload must run on instances optimized for sustained high memory usage relative to compute. Which instance family characteristic applies?

1. A A memory optimized family, which provides a higher ratio of memory to vCPU.
2. B A storage optimized family, which provides high local disk throughput.
3. C A general purpose family, which balances the two.
4. D A compute optimized family, which provides more vCPU per unit of memory.

150. A containerized workload must scale to zero when no requests arrive and start quickly when they resume. Which solution meets these requirements?

1. A Run the containers on a serverless container platform that scales to zero, accepting a brief start latency.
2. B Run the containers on an ECS service with a minimum of one task.
3. C Run the containers on a Dedicated Host.
4. D Run the containers on EC2 instances in an Auto Scaling group with a minimum of one.

151. A workload's instances must be placed to minimize the impact of a hardware failure. Which placement strategy applies?

1. A A partition placement group with a single partition.
2. B No placement group, which allows any placement.
3. C A spread placement group, which places instances on distinct underlying hardware.
4. D A cluster placement group, which packs instances closely together.

152. A relational workload must scale write throughput beyond what a single instance provides. Which approach applies?

1. A Increase the instance's provisioned IOPS.
2. B Shard the data across several database instances so writes are distributed.
3. C Enable Multi-AZ on the instance.
4. D Add read replicas to the existing instance.

153. A DynamoDB table must support a query pattern discovered after the table was created, using a different partition key. Which solution meets these requirements?

1. A Create a global secondary index with the required partition key.
2. B Create a local secondary index with the required attribute.
3. C Scan the table and filter on the attribute.
4. D Recreate the table with the new partition key.

154. An application requires a database that automatically scales storage and compute with no capacity planning for an intermittent workload. Which solution meets these requirements?

1. A Amazon RDS with a provisioned instance class.
2. B Amazon EC2 running a self-managed database.
3. C Amazon Redshift with a provisioned cluster.
4. D Amazon Aurora Serverless v2.

155. A workload must cache database query results with the cache surviving a node failure. Which solution meets these requirements?

1. A Amazon S3 used as a cache.
2. B Amazon ElastiCache for Memcached with several nodes.
3. C A local in-memory cache on each application instance.
4. D Amazon MemoryDB, which provides in-memory performance with durable multi-AZ storage.

156. A database migration must move a self-managed PostgreSQL database to Amazon RDS with minimal downtime. Which solution meets these requirements?

1. A Take a backup and restore it during a maintenance window.
2. B Recreate the schema and reload the data manually.
3. C Export the data to CSV and import it.
4. D Use AWS DMS with a full load followed by ongoing change data capture, then cut over.

157. An application's queries join data across tables that are always accessed together. Which DynamoDB design consideration applies?

1. A Increase the tables' provisioned read capacity.
2. B Create a global secondary index spanning both tables.
3. C Use a transaction to read both tables.
4. D Denormalize the related data into a single item or item collection, since DynamoDB does not perform joins.

158. An application must be reachable by a stable IP address that does not change if the backend is replaced. Which solution meets these requirements?

1. A Use an Application Load Balancer's DNS name.
2. B Use a Route 53 alias record.
3. C Use AWS Global Accelerator, which provides static anycast addresses fronting the backend.
4. D Assign an Elastic IP address to each backend instance.

159. Traffic between two VPCs must not traverse the internet and both VPCs are in different Regions. Which solution meets these requirements?

1. A Create a Site-to-Site VPN between the VPCs over the internet.
2. B Assign public IP addresses and restrict by security group.
3. C Create an inter-Region VPC peering connection or peer transit gateways across the Regions.
4. D Route the traffic through a NAT gateway in each VPC.

160. An application's users must reach the nearest of several Regional endpoints, with unhealthy endpoints removed. Which Route 53 configuration applies?

1. A Failover records with a primary and secondary.
2. B Geolocation records mapping countries to Regions.
3. C Latency records with a health check attached to each.
4. D Weighted records split by capacity.

161. A solution must run SQL transformations over data in Amazon S3 without managing servers. Which solution meets these requirements?

1. A Amazon Athena, which queries S3 data with SQL serverlessly.
2. B Amazon EMR with a persistent cluster.
3. C Amazon EC2 running a SQL engine.
4. D Amazon RDS with the data loaded into tables.

162. A pipeline must catalog the schema of data in Amazon S3 so several query engines can use it. Which solution meets these requirements?

1. A Define the schema separately in each query engine.
2. B Store the schema in an Amazon S3 object.
3. C Run an AWS Glue crawler to populate the Glue Data Catalog.
4. D Record the schema in a wiki page.

163. Streaming data must be transformed before delivery to Amazon S3 with no servers to manage. Which solution meets these requirements?

1. A Run a Spark Streaming job on an EMR cluster.
2. B Configure a Lambda transformation on an Amazon Data Firehose delivery stream.
3. C Transform the records in the producer application.
4. D Write the data to S3 and run a nightly batch transformation.

164. A solution must move data from an on-premises NFS share to Amazon S3 on a recurring schedule. Which solution meets these requirements?

1. A Copy the files with the AWS CLI when needed.
2. B Deploy AWS DataSync with a scheduled task from the NFS share to the bucket.
3. C Mount the bucket on the file server.
4. D Order AWS Snowball devices each month.

165. A solution must query data across Amazon S3 and Amazon Redshift in a single SQL statement. Which solution meets these requirements?

1. A Export the Redshift data to S3 and query with Athena.
2. B Use Amazon Redshift Spectrum to read the S3 data through an external schema.
3. C Load the S3 data into Redshift first.
4. D Query each source separately and combine the results.

166. A pipeline must process records as they arrive with sub-second latency and ordered processing per key. Which solution meets these requirements?

1. A Amazon S3 event notifications.
2. B A scheduled batch job.
3. C Amazon SQS standard queue.
4. D Amazon Kinesis Data Streams with the key as the partition key.

167. A solution must prepare data visually without writing code. Which solution meets these requirements?

1. A AWS Glue Spark jobs.
2. B AWS Glue DataBrew, which provides a visual data preparation interface.
3. C Amazon Athena queries.
4. D Amazon EMR notebooks.

168. A solution must convert incoming JSON records to Apache Parquet as they are delivered to Amazon S3. Which solution meets these requirements?

1. A Enable record format conversion on the Firehose delivery stream using a Glue Data Catalog schema.
2. B Convert the records in the producer.
3. C Convert the files with a scheduled Glue job after delivery.
4. D Store the JSON and convert at query time.

169. A solution must orchestrate a sequence of data processing steps with conditional branching. Which solution meets these requirements?

1. A A single Lambda function containing the logic.
2. B Amazon EventBridge rules triggered on a schedule.
3. C AWS Step Functions, which expresses sequencing and branching declaratively.
4. D Amazon SQS queues between each step.

170. A solution must search and analyze log data with full-text queries and dashboards. Which solution meets these requirements?

1. A Amazon Athena over the raw logs.
2. B Amazon Redshift with SQL queries.
3. C Amazon DynamoDB with a secondary index.
4. D Amazon OpenSearch Service, which provides full-text search and visualization.

171. A solution must process a large dataset with Apache Spark without managing cluster infrastructure. Which solution meets these requirements?

1. A Amazon EMR on EC2 with a persistent cluster.
2. B Amazon EC2 with Spark installed.
3. C Amazon Athena with a standard SQL query.
4. D AWS Glue, which runs Spark jobs serverlessly.