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

AWS DataSync provides managed, incremental, parallelized transfers between EFS file systems across Regions, supporting scheduled or continuously running tasks with automatic change detection, encryption in transit, and integrity verification. You can configure two tasks in opposite directions (bi-directional) to achieve two-way synchronization with high availability using agents in each Region and EFS locations connected via VPC endpoints. Native EFS replication (B) is one-way (read-only target) and not intended for active/active two-way sync. Options A and D rely on custom rsync and cross-Region mounts or SFTP, which introduce operational overhead, latency, and fail to provide resilient, managed synchronization. DataSync minimizes operational burden while delivering fault-tolerant, scalable, cross-Region EFS sync.

The requirements emphasize high availability across multiple Availability Zones, automatic recovery, and least operational overhead. The simplest way to achieve this for containerized microservices is to use a fully managed container orchestration service with serverless compute. Amazon ECS with the Fargate launch type provides exactly that.

With ECS Fargate, AWS manages the underlying compute infrastructure (no EC2 instances to provision, patch, scale, or secure at the OS level). You define task definitions and services, and ECS schedules tasks across Availability Zones based on your networking configuration (multiple subnets in different AZs). If a container fails, ECS service scheduling automatically replaces it, fulfilling the “recover from failures” requirement. Fargate also integrates with load balancing, CloudWatch logging, and scaling policies, reducing the operational burden of keeping the platform healthy.

Option A (ECS on EC2) is still managed orchestration, but it requires managing the EC2 capacity layer: AMI updates, instance patching, cluster capacity planning, and instance scaling. Option B (EKS with self-managed nodes) typically carries even more operational responsibility because you manage node groups, upgrades, and Kubernetes operational complexity in addition to the underlying instances. Option D is the highest operational overhead because it lacks managed scheduling and self-healing; you would need to build and maintain your own process manager and failover mechanisms.

Therefore, C is the best answer because it provides a managed, multi-AZ capable, self-healing container platform with minimal infrastructure management—matching the operational efficiency goal while maintaining high availability.

Amazon RDS Proxy helps manage and pool database connections from serverless compute like AWS Lambda, significantly reducing the stress on the database during unpredictable traffic surges. It improves scalability and resiliency by efficiently managing connections, protecting the database from being overwhelmed, and enabling failover handling.

Option A is the most cost-effective and operationally efficient approach to handling unpredictable surges and improving resilience without requiring major application changes.

Option B involves a migration to DynamoDB, which is a significant architectural change and costlier initially. Option C is manual connection cleanup, insufficient for unpredictable surges. Option D increases resources but does not solve connection storm problems efficiently and is more costly.

Aurora Global Database is designed for low-latency cross-Region reads and disaster recovery for relational data.

Amazon S3 Cross-Region Replication (CRR) automatically replicates unstructured data to another Region, ensuring high availability and resilience.

“Aurora Global Database replicates your data with typical latency of less than 1 second to secondary AWS Regions.”

“Amazon S3 Cross-Region Replication automatically replicates every object uploaded to your bucket to a destination bucket in another AWS Region.”

— Aurora Global Database

— S3 Cross-Region Replication

This combination meets the multi-Region, high availability, fault-tolerant requirement for both relational and unstructured data.

AWS WAF (Web Application Firewall) is designed to protect public-facing web applications from common web exploits and allows for deep inspection of HTTP and HTTPS requests. By enabling logging on AWS WAF, you can gain insights into traffic flow, request sources, and blocked requests, which improves both security visibility and posture. Integrating AWS WAF logging with Amazon Kinesis Data Firehose allows automatic, near-real-time delivery of log data to Amazon S3 for analysis, reporting, or integration with analytics tools. This solution not only secures the website but also enables comprehensive traffic analysis, satisfying both requirements efficiently.

Reference Extract from AWS Documentation / Study Guide:

" AWS WAF provides detailed logs of web requests, which can be delivered to Amazon S3 via Amazon Kinesis Data Firehose. This logging enables analysis of web traffic and sources, supporting both security and operational monitoring. "

Source: AWS Certified Solutions Architect – Official Study Guide, Security and Compliance section; AWS WAF Developer Guide (Logging and Monitoring).

AWS Secrets Manager is the recommended service for storing database credentials and performing automated rotation. Any Lambda function version or alias can fetch the latest secret value at runtime, ensuring no outdated credentials exist in deployed versions.

Environment variables (Option C) are static per version. Embedding credentials in code (Option B) is insecure and requires redeployment. Parameter Store (Option D) supports rotation but requires more configuration and is not as seamless as Secrets Manager for database credential rotation.

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

To ensure high availability and scalability, the web application should run in anAuto Scaling groupacross two Availability Zones behind anApplication Load Balancer (ALB). The database should be migrated toAmazon RDSwithMulti-AZ deployment, which ensures fault tolerance and automatic failover in case of an AZ failure. This setup minimizes administrative overhead while meeting the company ' s requirements for high availability and scalability.

Option A: Read replicas are typically used for scaling read operations, and Multi-AZ provides better availability for a transactional database.

Option B: Replicating across AWS Regions adds unnecessary complexity for a single web application.

Option D: EC2 instances across three Availability Zones add unnecessary complexity for this scenario.

AWS References:

Auto Scaling Groups

Amazon RDS Multi-AZ

AWS guidance for global, highly available, low-latency applications using a relational database recommends:

Amazon CloudFront with Amazon S3 for static content to cache data at edge locations globally and minimize latency for static assets.

A regional application tier deployed close to users using managed container services such as Amazon ECS on AWS Fargate, which removes the need to manage servers and scales automatically, reducing operational overhead.

A relational database tier using Amazon Aurora global database, which is purpose-built for globally distributed applications. Aurora global database provides a primary cluster in one Region with low-latency read replicas in secondary Regions and fast cross-Region replication, enabling low-latency reads and high availability for global users.

Option A uses only a single Region, which does not meet the “global users with low latency” requirement.

Option C uses RDS and DMS-based replication, which requires more management and does not provide Aurora’s integrated global database features.

Option D replaces the relational database with DynamoDB, which violates the requirement that the application interacts with a relational database.

Instances in a private subnet cannot directly reach the internet because they do not have public IP addresses and the private subnet route table typically does not send 0.0.0.0/0 traffic to an internet gateway. However, these instances still need outbound-only internet access to download patches and updates while remaining inaccessible from the internet. The standard AWS design to achieve this is a NAT gateway deployed in a public subnet.

Option C is required because a NAT gateway must reside in a public subnet that has a route to the internet gateway. This allows the NAT gateway to forward outbound traffic from private instances to the internet and return responses, without allowing inbound connections initiated from the internet to those instances.

Option A is required because a NAT gateway uses an Elastic IP address to represent its public-facing identity on the internet. Without an Elastic IP, the NAT gateway cannot communicate with internet endpoints. The Elastic IP is associated with the NAT gateway, not with the internet gateway.

Option B is required because the private subnet needs a route for 0.0.0.0/0 (default route) that targets the NAT gateway. This ensures that outbound internet-bound traffic from the private instances is sent to the NAT gateway rather than being dropped.

Option D is incorrect because a NAT gateway in a private subnet would not have a working route to the internet gateway and would not provide internet egress. Option E is incorrect because the public subnet’s default route should point to the internet gateway, not to the NAT gateway. Option F is incorrect because you do not associate Elastic IPs with an internet gateway.

Therefore, A, B, and C correctly implement private subnet outbound internet access while keeping the instances unreachable from the internet.

For applications requiring high network throughput and low latency between EC2 instances, AWS recommends using a Cluster Placement Group:

Cluster Placement Group: Packs instances close together inside an Availability Zone, enabling workloads to achieve the low-latency network performance necessary for tightly-coupled node-to-node communication.

Enhanced Networking: Selecting instance types that support enhanced networking provides higher bandwidth, higher packet-per-second (PPS) performance, and consistently lower inter-instance latencies.

By combining a Cluster Placement Group with enhanced networking-enabled instance types, the company can meet the stringent performance requirements of its real-time industrial application.

For short, frequent daily spikes (15–20 minutes), the most cost-effective scaling behavior typically comes from target tracking scaling driven by a request-based metric, because it continuously adjusts capacity to maintain a chosen target value rather than adding/removing capacity in coarse steps. Option D uses a custom CloudWatch metric representing the number of GET requests (sourced from ALB access logs, application instrumentation, or request counting logic) and applies target tracking so the Auto Scaling group scales out and in proportionally to demand. This helps avoid both under-provisioning (bad performance during spikes) and over-provisioning (wasted cost between spikes), which is exactly what cost-optimized elasticity aims for.

Option A (simple scaling) is usually less cost-efficient for bursty traffic because it relies on fixed adjustments and alarm thresholds, often with cooldowns. With 15–20 minute spikes, simple scaling can react too slowly, overshoot, or oscillate, leaving extra instances running after the spike or failing to ramp fast enough at the start. Option B (predictive scaling) is intended for regular, forecastable demand patterns, but it’s not always the most cost-effective choice for repeated short spikes because it can pre-scale conservatively and may require more historical data and tuning; it is often paired with dynamic scaling anyway. Option C uses UnhealthyHostCount, which is a health signal—not a demand signal—and scaling on unhealthy hosts can increase cost without addressing request volume; it may also mask underlying application issues rather than scaling to meet load.

Because the requirement explicitly allows standard or custom metrics and focuses on scaling based on request volume, D best matches a cost-efficient, responsive approach that directly tracks incoming workload.

The correct answer isBbecause the company needs tocentrally manage backups across multiple AWS accounts, improvedisaster recovery posture, and protect backups fromaccidental deletion or compromise of an AWS account.AWS Backup policies in AWS Organizationsprovide centralized governance and automation for backup plans across member accounts. By storing copies of backups inadditional AWS accounts, the company gains stronger isolation and protection against account-level compromise.

Cross-account backup copies are especially important when the requirement includes protection from a malicious attack on an AWS account. If backups remain only in the same account, an attacker with sufficient permissions in that account could potentially alter or delete them. Storing backup copies in a separate account provides administrative separation and strengthens recovery options. This is a recognized AWS best practice for improving cyber resilience and recovery assurance.

Option A is incorrect because copying backups to additional Availability Zones within the same account does not address centralized management across Organizations or protect against malicious actions in that account. Option C improves regional resilience but still lacks centralized policy-based management and account isolation. Option D provides centralized policy management and cross-Region resilience, but it does not directly address the requirement to protect against an attack on a single AWS account as effectively ascross-account backup copies.

AWS guidance for enterprise backup strategy emphasizescentralized backup management with AWS Backup and AWS Organizations, pluscross-account copiesfor stronger protection and recovery. Therefore, usingAWS Backup policies in Organizations to store copies in additional AWS accountsis the best solution.

Provisioned concurrency is the key requirement match because AWS documentation states that it pre-initializes Lambda execution environments and is designed to reduce cold-start latency, supporting predictable response times in the double-digit millisecond range. That makes it ideal for APIs that must remain consistently low-latency during peak traffic periods. Reserved concurrency does not solve cold starts; it only reserves capacity. ECS and EKS could certainly host the API, but both introduce more infrastructure management than API Gateway plus Lambda. Since the question asks for theleast operational overheadwhile maintaining consistent low latency, the serverless pattern with API Gateway and Lambdawith provisioned concurrencyis the strongest AWS-native answer. (AWS Documentation)

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

Comprehensive and Detailed Explanation (paraphrased from AWS Solutions Architect guidance):

Amazon Route 53 latency-based routing is specifically designed to route users to the endpoint that provides the lowest latency, based on actual latency measurements between AWS Regions and users’ networks. This directly supports the goal of optimizing load times (performance).

For non-AWS endpoints (like an on-premises data center), Route 53 lets you create a latency routing record and associate it with the AWS Region that is closest to that endpoint. In this case, the on-premises data center is near us-west-1, so you associate that latency record with the us-west-1 Region.

Users for whom the eu-central-1 Region has lower latency will be routed to the AWS website in eu-central-1.

Users for whom the us-west-1–associated endpoint has lower latency will be routed to the on-premises website.

This matches AWS best practices for high-performance, low-latency architectures when you have multiple endpoints in different geographic locations, including a mix of AWS and on-premises resources.

Why the other options are not correct:

Option A (Failover routing):

Failover routing is designed for high availability and disaster recovery (active–passive), not for optimizing latency.

All traffic goes to the primary endpoint (eu-central-1) unless it is unhealthy. That does not choose the lowest-latency endpoint for each user, so it does not meet the “optimize load times as much as possible” requirement.

Option B (IP-based routing):

IP-based / geo-based policies are about directing users based on location information (IP/geo), not measured latency.

You would have to manually decide which IP ranges get which endpoint, which does not guarantee that each user goes to the lowest-latency endpoint. It’s less precise and less dynamic than latency-based routing.

Option D (Weighted routing):

Weighted routing splits traffic according to fixed weights (e.g., 50/50), regardless of user location or latency.

This is useful for testing, blue/green deployment, or gradual traffic shifting, but not for performance optimization per user.

Therefore, the only option that directly aligns with AWS guidance for performance optimization and latency-based traffic steering across Regions and on-premises endpoints is:

✅ C. Create a DNS record with a latency-based routing policy… associate the on-premises endpoint with us-west-1.

Option Censures dynamic scaling based on demand using a target tracking scaling policy, optimizing costs.

Option Aresults in over-provisioning, leading to higher costs.

Option Bincreases costs by using larger instances.

Option Dis not feasible as instance store volumes are ephemeral and unsuitable for shared storage like EFS.

A VPC peering connection with proper route table entries provides private IP connectivity between EC2 instances in separate VPCs. This keeps database traffic off the public internet and satisfies the private communication requirement. To make the traffic secure in transit, the MySQL instance should use an SSL certificate so the application can establish encrypted database connections. RDS Proxy is not valid because the database is self-hosted on EC2, not Amazon RDS. Allowing a public IP address would violate the private traffic requirement. A network ACL also cannot reference a security group ID, so option E is not a valid control. The correct combination is private VPC-to-VPC routing plus database-level encryption in transit.

To grant cross-account access to an Amazon S3 bucket, you can use a bucket policy that specifies the AWS account ID of the account you want to grant access to. This method allows Account B to access the S3 bucket in Account A without the need for additional services or configurations.

Application Load Balancer (ALB) supports HTTP/HTTPS layer 7 routing, including header-based routing, path-based routing, and host-based routing.

AWS WAF is designed to provide rules-based filtering (block, allow, count) of HTTP(S) requests to protect applications from common exploits and malicious traffic.

ALB integrates directly with AWS WAF, so you can attach a web ACL with custom rules defined by the security team to block specific patterns while still using header-based routing.

Why the others are not correct:

A: Mutual TLS adds client certificate authentication but does not provide rules-based blocking or WAF-style inspection.

C: AWS Config is for configuration compliance and auditing, not request filtering.

D: NLB operates at layer 4 (TCP/UDP); it does not support HTTP header-based routing.

Amazon S3 Multi-Region Access Points allow applications to access S3 buckets in multiple Regions through a single global endpoint. This provides active-active read access with automatic routing to the closest bucket for low latency. With cross-Region replication, writes in the primary Region are automatically copied to the secondary Region, providing fault tolerance. Option A (CloudFront) provides caching and distribution, but does not address write replication or active-active bucket access. Option B (Transfer Acceleration) optimizes uploads across distances but does not enable cross-Region fault tolerance. Option C (AWS Backup) is designed for backup/restore, not real-time multi-Region reads and writes. Therefore, D is the correct solution for active-active read access and disaster recovery.