SAA-C03 Design High-Performing Architectures Practice Test 7 — 15 Questions

Question 1

A company needs to store and retrieve user-uploaded documents that are accessed infrequently but must be immediately available when requested. The documents are 5 MB on average and are accessed a few times per month. The solutions architect wants the most cost-effective storage solution that meets these requirements. Which AWS service should be used?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Amazon S3 Standard-Infrequent Access (S3 Standard-IA). S3 Standard-IA is the most cost-effective choice for infrequently accessed data that must be immediately available. It offers lower storage costs than S3 Standard while maintaining high durability and rapid access. The access pattern of a few times per month fits the infrequent access model, and there is no retrieval fee for S3 Standard-IA. Other options either have retrieval latency (Glacier Deep Archive) or reduced durability (One Zone-IA) or are not object storage (EBS).

Answer

Amazon S3 Glacier Deep Archive

Answer

Amazon S3 One Zone-Infrequent Access (S3 One Zone-IA)

Answer

Amazon Elastic Block Store (Amazon EBS) gp2 volumes

Question 2

A web application serves dynamic content from an Application Load Balancer (ALB) in front of Amazon EC2 instances. The application experiences sudden spikes in traffic, and the operations team wants to ensure the ALB can handle the load without manual intervention. The team also wants to minimize latency for users. What should a solutions architect recommend?

Accepted Answer

Enable cross-zone load balancing on the ALB and configure an Auto Scaling group for the EC2 instances.. Enabling cross-zone load balancing ensures even traffic distribution across all targets, and an Auto Scaling group automatically adds or removes EC2 instances based on demand. This combination provides elasticity and high availability, allowing the application to handle sudden traffic spikes without manual intervention. The ALB scales automatically, but backend capacity must be scaled, which Auto Scaling handles. Other options either do not scale the backend or introduce unnecessary changes.

Answer

Configure the ALB to use a larger instance type for the load balancer nodes.

Answer

Replace the ALB with a Network Load Balancer (NLB) and enable sticky sessions.

Answer

Enable AWS Global Accelerator and create an endpoint group for the ALB.

Question 3

A financial analytics company runs a read-heavy reporting application on an Amazon RDS for MySQL Multi-AZ DB instance. During month-end reporting, the primary instance CPU reaches 95% and report generation slows dramatically. The team wants to offload read traffic without changing the application's connection string or code, and they need the lowest operational overhead. Which solution meets these requirements?

Accepted Answer

Migrate to an Amazon Aurora MySQL cluster and use the reader endpoint, which automatically load-balances connections across Aurora Replicas.. Aurora MySQL's reader endpoint gives a stable, load-balanced DNS name that distributes read-only connections across Aurora Replicas, letting the reporting workload scale horizontally without application code changes. Read replicas would require endpoint changes, Multi-AZ standbys cannot serve reads, and vertical scaling does not isolate read traffic from writes.

Answer

Create a read replica and modify the application to use a separate reader endpoint.

Answer

Increase the primary DB instance class and enable storage autoscaling.

Answer

Enable Multi-AZ standby reads by modifying the DB instance parameter group.

Question 4

A company runs a batch processing job on Amazon EC2 instances that read and write large amounts of data to an Amazon EBS volume. The job requires high throughput and low latency. The instances are launched in a cluster placement group. The team wants to maximize EBS performance. Which EBS volume type should they use?

Accepted Answer

Provisioned IOPS SSD (io2 Block Express). For a batch processing job requiring high throughput and low latency, Provisioned IOPS SSD (io2 Block Express) is the best choice. It delivers the highest IOPS and throughput with sub-millisecond latency, and it supports cluster placement groups for enhanced network performance. Other EBS types either lack the necessary performance or are optimized for different use cases.

Answer

Throughput Optimized HDD (st1)

Answer

General Purpose SSD (gp3)

Answer

Cold HDD (sc1)

Question 5

A logistics company ingests vehicle telemetry into an Amazon Kinesis Data Stream with 8 shards. Each producer writes records using the vehicle ID as the partition key. During peak hours, a few high-volume vehicles cause hot shards, and consumers see increased iterator age. The team wants to improve throughput and reduce consumer lag with minimal application changes. Which action should they take?

Accepted Answer

Reshard the stream to add more shards and use enhanced fan-out for the consumers.. Resharding adds write capacity and dilutes hot partition keys, while enhanced fan-out removes the shared 2 MB/s per-shard read limit by giving each consumer dedicated throughput. Retention changes and random keys do not address both sides, and batch tuning cannot exceed the per-shard read ceiling.

Answer

Enable AWS Lambda event source mapping with a larger batch size for the consumers.

Answer

Increase the stream's retention period from 24 hours to 168 hours.

Answer

Switch the producers to use a random partition key instead of the vehicle ID.

Question 6

A financial analytics company ingests market data into an Amazon S3 bucket, then runs a nightly Spark job on an Amazon EMR cluster to aggregate the data. The job reads the same 2 TB S3 dataset multiple times per run, which increases job duration and S3 request costs. The team wants to improve the Spark job's read performance without changing the source data or the job logic. Which solution meets these requirements?

Accepted Answer

Configure the EMR cluster to use EMRFS with an Amazon S3-based cache.. The EMRFS S3 read cache keeps frequently accessed objects on the cluster's local instance store, so repeated Spark reads avoid repeated S3 GETs and network round trips. This directly addresses the repeated-read pattern without altering the source data or job code. Transfer Acceleration, a colder storage class, or manual EBS copies do not provide in-cluster read caching and would not reduce repeated S3 reads.

Answer

Change the S3 bucket to use S3 Standard-Infrequent Access storage class.

Answer

Enable Amazon S3 Transfer Acceleration on the source bucket.

Answer

Attach an Amazon EBS gp3 volume to each core node and copy the dataset there before the job.

Question 7

A logistics company runs a read-heavy reporting API backed by an Amazon RDS for MySQL Multi-AZ DB instance. Read replicas were added, but the application still sends all read queries to the primary endpoint, so replicas sit idle while the primary shows high CPU. The team wants to offload read traffic to the replicas without rewriting the application's connection logic. (Choose two.)

Accepted Answer

Point the application's read-only connection string at the reader endpoint of an Amazon RDS Proxy associated with the DB instance.. The application must keep its existing query code while read traffic moves off the primary. Both the RDS reader endpoint and an RDS Proxy reader endpoint accomplish this by presenting a single connection target that resolves to the replicas, so changing only the read-only connection string is sufficient. RDS Proxy additionally pools connections, while the reader endpoint is the simpler native option.

Answer

Enable Multi-AZ on each read replica so that reads fail over between replicas automatically during an Availability Zone event.

Answer

Add an Amazon ElastiCache for Redis cluster in front of the DB instance and cache the reporting query results there.

Answer

Increase the size of the primary DB instance class so it can serve both reads and writes with lower CPU utilization.

Question 8

A genomics research team runs a containerized variant-calling pipeline on Amazon ECS with the EC2 launch type. The pipeline reads a 4 TB reference genome from Amazon S3 at the start of every task, and this initial read dominates task startup time. The team wants to reduce startup latency for thousands of tasks per day while keeping the reference data highly durable and available. Which solution is MOST cost-effective?

Accepted Answer

Use an Amazon ECS-optimized Amazon Machine Image with a warm local cache and an S3 Gateway endpoint, or use a cached volume backed by Amazon S3 via AWS DataSync or a custom cache.. The bottleneck is the repeated 4 TB read from S3 for every task. Caching the reference genome on the container instance's local storage, while keeping S3 as the durable source of truth, removes that network read from the critical path. EFS, Transfer Acceleration, and scheduled DataSync copies either still transfer the full dataset or add operational complexity and cost, so local caching is the most cost-effective way to cut startup latency.

Answer

Create an Amazon S3 access point and enable S3 Transfer Acceleration for the bucket.

Answer

Store the reference genome on an Amazon EFS file system and mount it into each task.

Answer

Use AWS DataSync to copy the reference genome to each container instance's local storage on a schedule.

Question 9

A software company serves a containerized API to users across North America and Europe. The API runs on Amazon ECS tasks behind an Application Load Balancer in a single AWS Region. Users in Europe report high latency, and the team wants to improve global read performance while keeping a single authoritative database. Which two actions should the team take? (Choose two.)

Accepted Answer

Configure AWS Global Accelerator with an endpoint group pointing to the Application Load Balancer.. CloudFront caches cacheable API responses near European users, and Global Accelerator optimizes the network path for dynamic requests that must reach the origin. Together they reduce latency for both cached and uncached traffic while keeping one authoritative database. Regional compute without local data, S3 acceleration, and more tasks do not address the long-haul path.

Answer

Enable Amazon S3 Transfer Acceleration on the bucket that stores static API assets.

Answer

Deploy a second ECS cluster in a European Region and point it at the same single-Region database.

Answer

Increase the ECS task count in the existing Region to handle global traffic.

Question 10

A healthcare company runs a critical web application on Amazon EC2 instances behind an Application Load Balancer (ALB). The application stores session state in a self-managed Redis cluster on a single EC2 instance. During peak hours, users experience intermittent 5xx errors and slow response times. The operations team has observed that the Redis instance is CPU-bound and that the application servers are not the bottleneck. The company wants to improve performance and availability with minimal application changes. Which solution should a solutions architect recommend?

Accepted Answer

Replace the self-managed Redis instance with Amazon ElastiCache for Redis in cluster mode enabled.. The self-managed Redis instance is CPU-bound and a single point of failure. Amazon ElastiCache for Redis in cluster mode enabled offers a managed, highly available, and scalable Redis service that can distribute the load across multiple shards. It requires minimal changes if the application uses a cluster-aware Redis client, and it eliminates the operational overhead of managing Redis on EC2, directly improving performance and availability.

Answer

Increase the size of the Redis EC2 instance and add a read replica in a different Availability Zone.

Answer

Migrate the session state to Amazon DynamoDB with on-demand capacity and update the application to use the DynamoDB SDK.

Answer

Enable sticky sessions on the ALB and store session state in the local memory of each EC2 instance.

Question 11

A media analytics company runs a real-time ingestion pipeline. Producers write clickstream events to an Amazon Kinesis Data Streams stream with 8 shards. Consumers run on an Amazon EC2 Auto Scaling group. During peak hours, the stream's WriteProvisionedThroughputExceeded and ReadProvisionedThroughputExceeded metrics increase, and consumers see elevated GetRecords latency. The team wants to improve throughput and reduce consumer latency. Which two actions should a solutions architect take? (Choose two.)

Accepted Answer

Enable enhanced fan-out for the consumers using the Kinesis Data Streams API.. WriteProvisionedThroughputExceeded means the stream's shards cannot absorb the ingest rate, so adding shards raises total write capacity and enables more parallelism. ReadProvisionedThroughputExceeded with multiple consumers means the shared 2 MB/sec per-shard read limit is the constraint, and enhanced fan-out gives each consumer dedicated throughput. Firehose changes the architecture, extra consumers beyond shard count cannot read more, and encryption does not affect throughput.

Answer

Enable server-side encryption on the Kinesis data stream with an AWS KMS customer managed key.

Answer

Switch the stream to Kinesis Data Firehose delivery to Amazon S3.

Answer

Increase the EC2 Auto Scaling group's maximum size to add more consumer instances.

Question 12

An Aurora PostgreSQL application has an OLTP writer and a reporting dashboard that issues many read-only queries. The writer is healthy, but read latency rises noticeably during reporting windows. Which two changes should you make? Select two.

Accepted Answer

Add Aurora Replicas to scale out the read workload.. Adding Aurora Replicas (Option A) is correct because Aurora Replicas are dedicated read-only instances that share the same underlying storage volume as the writer, allowing you to scale read capacity linearly without impacting write performance. Sending read-only traffic to the reader endpoint (Option B) is correct because the reader endpoint automatically load-balances connections across all available Aurora Replicas, ensuring that dashboard queries are distributed and do not overload a single instance.

Answer

Scale up only the writer instance and keep all queries on it.

Answer

Replace the cluster with a single-AZ RDS instance to reduce replication overhead.

Answer

Move the dashboard to DynamoDB without changing the query model.

Question 13

A production application writes to an Amazon Aurora PostgreSQL cluster. Users report that during business-hour reporting runs, write latency increases. The application team wants to keep the writer focused on OLTP writes while still providing low-latency reads for reporting queries. What architectural approach should the solutions architect recommend?

Accepted Answer

Create Aurora read replicas and direct reporting read-only connections to the cluster reader endpoint.. A is correct because creating Aurora read replicas and directing reporting read-only connections to the cluster reader endpoint offloads read traffic from the writer instance. This allows the writer to focus on OLTP writes, while the reader endpoint load-balances read-only queries across replicas, providing low-latency reads for reporting without impacting write performance.

Answer

Resize the writer instance to a larger class so it can handle both writes and reads with fewer slowdowns.

Answer

Enable cross-region replication for the entire cluster so reporting always runs in the secondary Region.

Answer

Disable read replicas and use caching only in the application layer, keeping all queries connected to the writer endpoint.

Question 14

A DynamoDB table stores device status items. The partition key is deviceId, and the partition distribution is healthy (no single partition dominates). However, during peak periods the application experiences high read latency because many clients repeatedly request the latest status for the same devices. Which action best improves read latency without changing the DynamoDB partitioning model?

Accepted Answer

Add Amazon DAX as a caching layer in front of DynamoDB and route repeated read operations through DAX.. Amazon DAX is a fully managed, in-memory cache for DynamoDB that provides microsecond read latency. By caching the results of repeated GetItem and Query requests for the same device status items, DAX offloads read traffic from the underlying DynamoDB table, reducing the number of read capacity units consumed and eliminating the latency caused by repeated fetches from disk. This directly addresses the high read latency during peak periods without altering the existing partition key or partitioning model.

Answer

Change the partition key to a random value for each request to eliminate hot partitions.

Answer

Increase write capacity only, because writes generally determine read latency in DynamoDB.

Answer

Create an additional Global Secondary Index (GSI) and read exclusively from the index to accelerate reads.

Question 15

An API team runs an AWS Lambda function behind an Application Load Balancer (ALB). During predictable hourly traffic spikes, p95 response latency increases due to occasional cold starts. The team wants stable latency during those spikes without permanently overprovisioning resources for all functions. Which configuration is the most appropriate way to reduce cold starts for this Lambda function?

Accepted Answer

Publish a version of the function and configure provisioned concurrency on an alias, using autoscaling for the alias.. Provisioned concurrency initializes a specified number of execution environments in advance, keeping them warm and ready to handle requests without cold start latency. By configuring provisioned concurrency on an alias with autoscaling, the team can dynamically adjust the number of pre-warmed environments to match predictable traffic spikes, avoiding permanent overprovisioning while ensuring stable p95 latency.

Answer

Increase the function memory size and rely on faster initialization to reduce cold starts.

Answer

Set reserved concurrency equal to the expected peak requests per second for the function.

Answer

Use an event source mapping with a higher batch size so Lambda triggers earlier and keeps the runtime warm.