Amazon Web Services Updated SAP-C02 Exam Questions and Answers by boris

A is correct:Quota request templates must be created in us-east-1, the only region that supports them. You specify the desired quotain the correct target Region(ap-southeast-2) for SageMaker training and endpoint usage. This enablesautomatic quota increasesfor newly created accounts in the org.

B, C, and D misconfigure either region or quota type.

The company wants three things: minimize client-to-broker latency within the same Availability Zone, increase available bandwidth, and increase the scaling speed of the MSK cluster. The current brokers are Standard brokers (two per AZ). Clients use default settings, which means they are not explicitly configured for rack awareness or AZ affinity.

A common way to reduce latency in multi-AZ Kafka deployments is to enable rack awareness on clients and brokers so clients prefer brokers in the same “rack,” which can map to an Availability Zone. In Kafka, the client.rack setting allows the client to include rack information so the broker can return metadata that helps the client select replicas that are closest, reducing cross-AZ traffic and improving latency.

To increase bandwidth and improve scaling speed, the most direct approach in the choices is to move from Standard brokers to Express brokers. Express brokers are designed to provide higher throughput and faster scaling characteristics compared to standard broker types. Since the question explicitly calls out increasing available bandwidth and scaling speed, the broker type change is the key lever, and it can be combined with client.rack configuration to minimize cross-AZ latency.

Option C matches these requirements: it replaces Standard brokers with Express brokers (to improve throughput/bandwidth and scaling speed) and sets client.rack to the Availability Zone identifier (az_id) to improve locality and reduce latency between clients and brokers in the same AZ.

Option A is not appropriate because MSK does not use EC2 Auto Scaling predictive scaling in that manner, and Kafka clients/brokers are not “associated” with an EC2 placement group as a primary latency solution in MSK. Placement groups are for EC2 instance placement; MSK broker placement is managed by the service.

Option B introduces a proxy layer and MSK Connect in a way that increases complexity and does not directly guarantee lower latency or higher bandwidth. MSK Connect is for Kafka Connect workloads, not as a general-purpose low-latency routing proxy for Kafka clients. Cruise Control is used for partition rebalancing and cluster optimization, but it does not replace the benefits of higher-throughput broker types and client rack awareness for AZ locality.

Option D increases broker size and introduces PrivateLink endpoints. PrivateLink is about private connectivity from VPCs to services and does not inherently ensure AZ-local broker selection or reduce latency between clients and brokers in the same AZ. Also, resizing to xlarge increases capacity but does not address scaling speed and locality as directly as express brokers plus rack configuration.

Therefore, option C best meets all requirements.

This solution will allow the detection logic to be run as soon as the image is uploaded to the S3 bucket, before it is served to users via the CloudFront distribution. This way, the detection logic can quickly identify any corrupted images and prevent them from being served to users, minimizing latency between ingestion and serving.

D is required because the only reliable way to ensure newly launched Auto Scaling instances are patched is to make the launch template reference an AMI that already includes the latest security updates (an immutable image approach). AWS Systems Manager can automate building and maintaining patched AMIs (for example, through automated image creation workflows), after which the launch template is updated to the new AMI and the fleet is updated using Instance Refresh. Instance Refresh performs a controlled rolling replacement of instances so that the Auto Scaling group converges to the new AMI baseline.

C complements D by ensuring safe replacement and availability during the refresh/replacement process. Placing an ALB in front of the Auto Scaling group with health checks ensures that only healthy, fully bootstrapped/patched instances receive traffic, and that traffic is drained away from instances being replaced. Monitoring target health confirms the rollout is successful and minimizes risk during patch-driven reboots or instance replacement.

Why the other options are incorrect:

A: A termination policy setting does not ensure new instances are patched. It only affects which instances are terminated first. It does not solve the “launch patched instances” requirement.

B: Running two Auto Scaling groups and continuing in-place patching increases operational overhead and still risks drift and unpatched capacity when scaling occurs outside the maintenance window. It also does not address the core issue: the launch template AMI baseline.

E: NLB + termination protection does not ensure instances are patched at launch. Termination protection can interfere with Auto Scaling’s ability to replace instances, and NLB does not inherently provide the same application-layer health check behavior and deployment safety patterns typically used for rolling replacements (compared to ALB target group health checks).