Amazon Web Services SAA-C03 - AWS Certified Solutions Architect - Associate (SAA-C03)

AWS documentation states that workloads running 24/7 should use Reserved Instances or Savings Plans for the lowest cost. Therefore, the frontend nodes, which always run, should use Reserved Instances.

The backend nodes process asynchronous, restartable jobs, which makes them ideal for EC2 Spot Instances, the most cost-effective compute option for interruption-tolerant workloads.

Fargate (Options A and D) is significantly more expensive for large compute usage. Spot Instances cannot be used for critical frontend nodes (Option C).

=====================================================

Versioning in S3 preserves every version of every object — preventing permanent overwrites. This ensures compliance where data cannot be overwritten or lost.

SSE-S3 ensures server-side encryption.

“When you enable versioning, Amazon S3 stores every version of every object. With versioning, you can preserve, retrieve, and restore every version of every object stored in an S3 bucket.”

— S3 Versioning

This satisfies regulatory requirements for protecting data from overwriting indefinitely.

Incorrect Options:

B: Object Lock with retention period is time-bound, not indefinite.

C: Legal hold blocks deletion but doesn’t directly prevent overwriting.

D: Storage Lens is analytics, not protection.

The design already uses one transit gateway (TGW) per Region and inter-Region TGW peering, which addresses the multi-Region AWS-side routing. The remaining requirement is to extend on-premises connectivity over Direct Connect so that on-premises networks can reach VPCs attached to the TGWs across all three Regions. The most operationally efficient and scalable approach is to use AWS Direct Connect Gateway (DXGW) with a transit virtual interface (transit VIF) and then associate the Regional transit gateways to that DXGW.

Option C is purpose-built for this: a transit VIF is specifically used to connect a Direct Connect connection to a Direct Connect gateway, and a DXGW can then be associated to multiple transit gateways (and can be used across Regions), enabling centralized connectivity from on-premises to TGW-connected VPCs. With TGW attachments and TGW route tables, you can propagate and control routes across many VPCs and accounts, which fits the “50 accounts, thousands of VPCs” scale. Inter-Region TGW peering then allows on-premises routes learned via the DXGW/TGW in one Region to reach workloads in the other Regions through the TGW peering relationships, subject to routing configuration.

Option A is too limited and not scalable because it ties Direct Connect to a virtual private gateway (VGW) associated with a single VPC, which does not meet the multi-VPC, multi-account hub-and-spoke requirement. Option B incorrectly suggests associating a DXGW with a VGW “in each VPC” (VGW is per VPC and would not scale well here, and it doesn’t integrate with the TGW hub design you’ve already built). Option D is not the intended pattern: Site-to-Site VPN and public VIF do not replace the DXGW + transit VIF architecture for large-scale TGW-based private routing.

Using a Lambda function as part of the Kinesis Data Firehose pipeline allows for real-time detection and masking of PII before data is written to S3. This ensures that PII is never stored in its raw form in the data lake.

Option B:Amazon Macie can scan and classify data but does not provide in-line PII masking for data ingestion.

Option C:Server-side encryption secures data but does not mask PII.

Option D:CloudHSM is unnecessary for PII masking and adds complexity without addressing the requirements.

AWS Documentation References:

Using Lambda with Kinesis Data Firehose

The problem is contention between two very different workloads that currently share one target group and one scaling boundary. The correct fix is to useALB path-based routingso /reports and /generateToken go toseparate target groups, each backed by its own Auto Scaling group. AWS ALB listener rules support path-based routing, which lets different paths scale independently and isolates the latency-sensitive /generateToken path from the CPU-heavy /reports workload. Making the instances larger does not remove the contention boundary, and CloudFront is not the right answer for two dynamic uncached paths. Therefore, splitting the paths into independent target groups and scaling units is the best architecture.

AWS best practice is to use IAM roles with instance profiles for EC2 instances so that applications obtain temporary credentials automatically and do not need to store access keys.

To honor the principle of least privilege, each microservice should have an IAM role that grants only the specific permissions it needs.

Therefore, creating individual IAM roles per microservice and attaching them via instance profiles (Option D) both minimizes long-term credential management and applies least privilege cleanly.

Why others are not correct:

A: Using IAM users with access keys in code is insecure and high-overhead (key rotation, secret management).

B: A single broad role violates least privilege because every microservice gets more permissions than it needs.

C: Separate accounts per microservice is extreme over-segmentation and significantly increases operational complexity.

An S3 Bucket Key reduces the cost of using AWS KMS for server-side encryption by decreasing the number of requests made to KMS. By enabling S3 Bucket Key, the company can meet its encryption requirements with KMS keys while optimizing costs by reducing the number of KMS API requests.

Key AWS features:

Cost Optimization: S3 Bucket Keys reduce the frequency of KMS calls, optimizing the cost associated with encryption while still using AWS KMS for key management.

Compliance with KMS Encryption: This solution continues to meet the strict encryption requirements of the company by using KMS managed keys.

AWS Documentation: Using an S3 Bucket Key is recommended for organizations looking to optimize encryption costs without compromising security​​.

To meet the requirements, the key must support automatic rotation, and the company must be able to disable (deactivate) the key and schedule deletion. In AWS KMS, these lifecycle controls are available for customer managed keys. Automatic key rotation is supported for symmetric customer managed KMS keys used for encryption and decryption operations. When automatic rotation is enabled, AWS KMS generates new cryptographic material for the key on a regular schedule while the key ID remains the same, helping organizations meet compliance and security best practices without manual operational work.

Option C is the only choice that explicitly combines a symmetric customer managed key with automatic rotation enabled, which directly satisfies the rotation requirement. As a customer managed key, it can also be disabled (to prevent use) and scheduled for deletion (with a waiting period), meeting the rest of the lifecycle needs.

Option A is incorrect because automatic rotation is not generally available for asymmetric KMS keys in the same way; asymmetric keys are used for specialized use cases such as signing or asymmetric encryption, and rotation behavior differs. Options B and D mention “disabling envelope encryption,” which is not a configurable “option” you turn off in KMS; envelope encryption is a recommended pattern where KMS protects data keys, and services commonly use it under the hood. Those options therefore do not describe a valid configuration that meets the requirement.

Therefore, the correct solution is to create a symmetric customer managed KMS key and enable automatic rotation, while using standard KMS controls to disable and schedule deletion when required.

Creating a listener rule on theApplication Load Balancer (ALB)to return a maintenance response during the maintenance window is the most straightforward solution with the least operational overhead. The rule can be configured to match all incoming requests and return a custom response, and it can be easily removed once maintenance is complete.

Option A (Aurora table flag): This adds unnecessary complexity for a temporary maintenance response.

Option B and D (SQS or SNS): These options introduce more components than needed for a simple maintenance message.

AWS References:

ALB Listener Rules

Amazon EFS is the correct answer because it provides aPOSIX-compliant, shared file systemthat multiple ECS tasks and instances can mount concurrently. AWS documents EFS as a managed NFS file system built for shared access and elastic growth, which fits the requirement for log and configuration sharing across Linux containers.Elastic throughputis the best cost-conscious choice when access patterns can vary, because throughput scales automatically without manual provisioning. S3 is object storage and does not provide native POSIX file system behavior. EBS is block storage and is not the right answer for shared, highly available multi-container access across a distributed environment. FSx for Lustre is optimized for specialized high-performance workloads, not general shared configuration and log storage.

============

AWS STS is the best choice because it issuestemporary security credentialsthat expire automatically, which removes the need to create and later deactivate long-lived IAM users. AWS guidance for third-party and short-term access explicitly recommends roles and temporary credentials instead of sharing long-term credentials. Using predefined IAM roles also makes it easy to assign different permission levels to different contractors while keeping access centrally controlled. IAM users create more administrative effort and leave credential-lifecycle management to the company. AWS Config and CloudTrail are useful for governance and auditing, but they are not the primary mechanism for granting short-term role-based access. Therefore, STS with predefined roles is the most secure and lowest-overhead solution.

============

For latency-sensitive, globally distributed applications such as online gaming, minimizing network latency between users and application endpoints is critical. AWS Global Accelerator is designed specifically for this purpose. It uses the AWS global network and Anycast IP addresses to route user traffic to the closest healthy endpoint, reducing latency and improving performance for users worldwide.

Option C is the best solution because Global Accelerator directs traffic over the AWS backbone network instead of the public internet, which significantly reduces jitter and latency. It also provides built-in health checks and automatic failover, improving availability as the service expands globally. Importantly, Global Accelerator works seamlessly with existing Application Load Balancers, allowing the company to enhance performance without redesigning the application stack.

Option A (CloudFront) is optimized for caching static and cacheable content and is not ideal for real-time, stateful gaming traffic. Option B adds unnecessary API abstraction and additional latency. Option D introduces regional duplication and DNS-based routing, which is slower to react to network conditions and does not provide the same low-latency routing as Global Accelerator.

Therefore, C meets the requirement for global expansion with the least possible latency while maintaining a simple and highly performant architecture.

Amazon RDS Proxy is a fully managed database proxy for RDS that makes applications more scalable, more resilient to database failures, and more secure. RDS Proxy pools and shares database connections, allowing applications to open and close connections as needed without overwhelming the database. This is the recommended solution when the application cannot be modified to use connection pooling itself.

AWS Documentation Extract:

" Amazon RDS Proxy helps manage database connections to improve application scalability and performance. It pools connections and shares them among application clients, which can mitigate issues caused by opening and closing many database connections. "

(Source: Amazon RDS Proxy documentation)

A: Upgrading the instance does not solve connection inefficiency and can be cost-ineffective.

B: Increasing IOPS only helps if storage is a bottleneck, not if connections are the issue.

D: Multi-AZ improves availability, not connection management.

Amazon QLDB is the most cost-effective and suitable service for maintainingimmutable, cryptographically verifiable logsof data changes. QLDB provides a fully managed ledgerdatabase with a built-in cryptographic hash chain, making it ideal for recording changes to accounting records, ensuring data integrity and security.

QLDB reduces operational overhead by offering fully managed services, so there’s no need for server management, and it’s built specifically to ensure immutability and verifiability, making it the best fit for the given requirements.

Option A (Redshift): Redshift is designed for analytics and not for immutable, cryptographically verifiable logs.

Option B (Neptune): Neptune is a graph database, which is not suitable for this use case.

Option C (Timestream): Timestream is a time series database optimized for time-stamped data, but it does not provide immutable or cryptographically verifiable logs.

AWS References:

Amazon QLDB

How QLDB Works

For logs that are:

Written and processed within a short period (24 hours)

Accessed quickly for compute/analytics

With unknown object count and size

Amazon S3 Standard is the most appropriate and cost-effective. Intelligent-Tiering is designed for data stored for longer periods (typically 30+ days) with changing access patterns and charges a per-object monitoring and automation fee that becomes inefficient for very short-lived objects.

S3 Glacier Deep Archive and S3 One Zone-IA are optimized for long-term archival or infrequently accessed data with retrieval time or availability constraints that are not suitable for nightly active processing.

Because the application is public and serves global customers with dynamic IP addresses, the web tier must allow HTTPS traffic from0.0.0.0/0on port 443. For the database tier, the secure design is to allow MySQL access only from theweb tier’s security group, not from the internet. AWS RDS documentation explains that you can specify another VPC security group as the source for the DB security group, which allows application servers to connect without exposing the database publicly. That makes A the correct configuration. Options B and C do not work well for global dynamic users, and D exposes the database to the internet.

============

Using Amazon CloudFront in front of an ALB in a single region is a cost-effective way to deliver content with low latency across the globe. CloudFront caches content closer to the users, reducing the load on backend servers and minimizing data transfer costs by serving cached content from edge locations.

Compared to Global Accelerator, CloudFront is significantly more cost-optimized for static and dynamic content delivery. Multi-region deployments increase infrastructure and transfer costs, which violates the optimization goal. Therefore, option A provides the best mix of performance and cost efficiency.

Amazon S3 Intelligent-Tiering is designed for unknown or unpredictable access patterns, automatically moving objects to the most cost-effective tier without performance impact.

It is ideal when you do not know object size or access frequency.

S3 Standard (Option B) works but costs more.

S3 One Zone-IA and Glacier Deep Archive (Options D and A) are not appropriate because data is frequently accessed for processing and persists only 24 hours.

=====================================================

AWSSecrets Manageris the best solution for securely storing and managing sensitive information, such as usernames and passwords. Secrets Manager provides automatic rotation, fine-grained access control, and encryption of credentials. It is designed to integrate easily with other AWS services, such as CloudFormation, to automate the retrieval of secrets via theBatchGetSecretValue API.

Secrets Manager has a lower operational overhead than manually managing credentials, and it offers features like automatic secret rotation that reduce the need for human intervention.

Option A and C (Parameter Store): While Systems Manager Parameter Store can store secrets, Secrets Manager provides more specialized capabilities for securely managing and rotating credentials with less operational overhead.

Option B and C (AWS Batch): Introducing AWS Batch unnecessarily complicates the solution. Secrets Manager already provides simple API calls for retrieving secrets without needing an additional service.

AWS References:

AWS Secrets Manager

Secrets Manager with CloudFormation