Amazon AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 52

Question 510

A T3 application server is configured in Unlimited mode. It holds CPU utilization above its baseline for many consecutive hours without the throttling seen in Standard mode, but the bill now contains CPU credit charges. Why?

Unlimited mode makes surplus CPU credits permanently free once the original accrued balance reaches zero.
Above-baseline sustained CPU can consume surplus credits that incur charges when usage exceeds the applicable credit earning and grace economics.
EBS gp3 IOPS are billed under the name CPU credits whenever an instance opens a large file.
An ASG health check increases the CPU credit balance on every instance without limits or price impact.

Explanation

Correct Answer

With Unlimited, usage above baseline can draw surplus credits and incur the documented additional charge under the applicable averaging rules. Compare sustained workload price/performance with a nonburstable instance before assuming the T-family is the economical fit.

Incorrect Answers

The Unlimited mode permits prolonged bursts, but its billing treatment is not identical to free baseline entitlement. Sustained excess CPU can be charged.

EBS I/O charges and EC2 CPU-credit charges represent different metering dimensions. Disk IOPS does not relabel itself as compute surplus credits.

Health checks can replace unhealthy fleet members, not mint free credits. The Unlimited-mode surplus accounting is driven by the instance CPU use and credit rules.

Question 511

An ALB-based Auto Scaling group uses a target-tracking ALBRequestCountPerTarget policy. New instances launch, but instance boot takes 4 minutes and the metric rises during that warm-up. The operator wants to prevent newly launched capacity being treated as immediately available. Which Auto Scaling consideration fits?

Use the total ALB RequestCount as the target metric because total requests decrease automatically when each new instance launches.
Set minimum desired instances to zero and assume requests can be served before a new instance passes its health checks.
Configure an appropriate instance warm-up/health-readiness window and verify targets become healthy before counting the capacity as fully productive.
Rename the target group to make booting instances instantly become healthy in the load balancer.

Explanation

Correct Answer

A new instance is not necessarily ready to handle normal traffic at launch. Effective warm-up/health checks and startup timing keep scaling decisions aligned to actual serving capacity instead of overestimating booting hosts.

Incorrect Answers

ALB total request volume can stay unchanged when targets scale; target-tracking request-per-target is the relevant proportional signal, and switching to a total metric does not solve warm-up accounting.

Reducing capacity to zero makes cold service initialization more important, not less. Traffic readiness still depends on successful health checks and startup.

Names do not alter target health or operating-system boot time. Readiness comes from launching the application and passing ALB target checks.

Question 512

An EC2 Auto Scaling predictive policy displays sensible hourly traffic forecasts in CloudWatch, but the group never launches instances ahead of predictable peaks. Policy Mode is set to ForecastOnly. What change addresses the missing pre-scaling?

After validating forecast accuracy and capacity bounds, change predictive policy mode to ForecastAndScale so its forecast can drive scaling actions.
Increase S3 lifecycle expiry to make a forecast-only policy launch EC2 instances even though its scaling mode is disabled.
Use a Route 53 geolocation record because DNS location policy changes predictive forecast mode automatically.
Set the Auto Scaling group minimum to zero and assume this changes ForecastOnly into ForecastAndScale.

Explanation

Correct Answer

ForecastOnly creates forecasts for observation but does not apply predictive capacity changes. ForecastAndScale enables the forecast to direct pre-scaling; monitor it and retain responsive policies for deviations.

Incorrect Answers

S3 object-retention settings are unrelated to EC2 predictive scaling policy mode and do not cause forecasted capacity launches.

A DNS routing policy has no authority to modify the Auto Scaling predictive policy or determine when the group launches new instances.

Group minimum capacity constrains a possible group size but is not a policy-mode switch. ForecastOnly still does not act on its predictions.

Question 513

A team schedules a 08:00 local-time EC2 Auto Scaling scale-out but observes capacity appearing after the local traffic surge. The action was saved in UTC and instance initialization takes another 7 minutes. What correction best addresses the timing error?

Specify the intended scheduling time zone (or correct UTC conversion) and move the action early enough for measured startup and target-health readiness.
Switch to the total S3 BucketSizeBytes metric because stored bytes define the time zone used for ASG scheduled actions.
Rename the scheduled action to include the local city; action names alone change how AWS interprets its stored UTC trigger.
Change Route 53 DNS TTL to zero and assume new EC2 instances are automatically booted seven minutes before traffic arrives.

Explanation

Correct Answer

Scheduled actions execute at configured times, not at the operator’s unstated wall-clock assumption. Correct time-zone conversion and startup lead time address both late trigger and warm-up.

Incorrect Answers

S3 storage inventory has no relationship to the execution timestamp of an Auto Scaling scheduled event.

Human-readable action names do not override the scheduled time and time-zone configuration. An actual update to the schedule is required.

DNS TTL affects resolution caching, not launch schedules or instance bootstrap. It cannot create capacity ahead of the configured trigger.

Question 514

A payment application keeps its only stateful database primary on a single EC2 Spot Instance, with no replication or failover. AWS reclaims that Spot capacity and processing stops. Which assumption was unsafe?

Spot capacity automatically reserves the instance through the entire selected month as long as its EC2 status checks pass.
A sole noninterruptible stateful primary must not rely on reclaimable Spot capacity without an availability and recovery architecture.
An S3 bucket versioning flag guarantees every in-memory payment transaction survives interruption of the sole database primary.
A Route 53 health check alone restores the lost database writes from memory onto a replacement host.

Explanation

Correct Answer

Spot is appropriate for interruptible/restartable work, not an unprotected unique transactional primary. A robust stateful tier needs durable writes, standby/failover and a capacity model meeting its interruption tolerance.

Incorrect Answers

Spot is spare capacity and may be interrupted when AWS needs it; health checking does not change its reclamation contract.

S3 object history can protect stored objects, not automatically mirror outstanding database state on an EC2 host that disappears.

DNS checks can route away from failed endpoints but do not create a new consistent copy of a lost transactional database.

Question 515

A mixed-instance EC2 Auto Scaling group is configured for a stable On-Demand base plus Spot burst capacity. Yet Spot scale-out fails at peak because the team limited allocation to one scarce instance type in a single Availability Zone. What adjustment best improves its chance of acquiring Spot capacity?

Remove the On-Demand base entirely so any shortage of the selected Spot pool has no effect on service availability.
Make every request larger than the biggest instance type in that pool, assuming capacity availability increases with oversized requests.
Expand eligible instance families/sizes and Availability Zones and use a capacity-optimized or price-capacity-optimized Spot allocation strategy while preserving the critical On-Demand baseline.
Disable instance health checks and assume one narrow Spot pool then becomes guaranteed reserved capacity.

Explanation

Correct Answer

Spot capacity varies across pools. Diversifying eligible instance types and AZs increases allocation opportunities, while an allocation strategy and On-Demand baseline balance cost and availability.

Incorrect Answers

Removing the stable baseline increases service exposure to Spot scarcity; it does not add supply to a constrained pool.

Increasing per-instance capacity demands may further restrict eligible hosts and does not broaden available Spot pools.

Health checks detect fleet health but cannot reserve or manufacture otherwise unavailable Spot capacity.

Question 516

An ECS service deploys containers that need 80 seconds to initialize. The ALB checks their target health immediately and repeatedly reports unhealthy; the ECS scheduler stops and replaces them before initialization completes. The application logs show no actual crash. What is the focused fix?

Increase the container task CPU limit to any value and assume that alone changes the ALB check to start after 80 seconds.
Change the S3 bucket retention policy so an ALB treats an initializing ECS container as healthy before HTTP readiness works.
Set a suitable ECS service health-check grace period and align ALB thresholds with realistic startup/readiness before concluding tasks have failed.
Rename the ECS task family because the scheduler gives longer health grace periods to certain family-name prefixes.

Explanation

Correct Answer

ECS can temporarily ignore relevant unhealthy checks after task start for a configured grace interval. Together with realistic ALB readiness thresholds, it lets valid slow-starting tasks become healthy before replacement.

Incorrect Answers

CPU capacity can affect startup if measured as the cause, but it does not by itself configure when health checks are used to stop tasks. This case identifies premature health-based replacement.

S3 retention has no control over ALB target readiness or ECS service health-check grace.

Task-definition names are identifiers, not health-check timing policy. The scheduler uses explicit service and load-balancer settings.

Question 517

An EKS pod requests 64 GiB of memory and remains Pending. The node autoscaler can add any number of workers, but every configured eligible node group uses instances with only 32 GiB RAM per node. What must change to make the pod schedulable?

Double the number of 32-GiB workers and assume the Kubernetes scheduler combines unused RAM from several separate machines for one ordinary pod.
Reduce the Auto Scaling scan interval to zero and assume a larger single-node memory requirement can be placed on existing 32-GiB shapes.
Reduce the pod memory request where technically safe or add an eligible node group with sufficient allocatable memory on a single node.
Change the cluster name so Kubernetes increases the RAM available on each existing EC2 instance without changing its type.

Explanation

Correct Answer

The scheduler needs a suitable single target node after system reservations. Adjust honest resource requests or provide sufficiently large nodes so the workload can be placed.

Incorrect Answers

A pod normally must fit on one node; node memory is not pooled as one schedulable machine for a single pod request.

Faster scheduling loops do not expand the actual per-node allocatable memory needed for this pod.

Names do not resize EC2 memory. Instance-type and resource-request changes are the relevant dimensions.

Question 518

A Lambda image-processing function has substantial allocated memory and low measured CPU utilization. Its duration is dominated by waiting on a fixed-latency external database over the network. Increasing configured memory and CPU entitlement does not materially shorten requests. What should be optimized?

Raise the function memory repeatedly because CPU sizing always removes every external network wait regardless of backend behavior.
Investigate database query/connection latency, caching and request fan-out; more Lambda CPU alone cannot eliminate the measured backend wait.
Change the Route 53 hosted-zone comment because the comment string controls all database query plans and Lambda execution time.
Increase Lambda maximum timeout and assume that granting a longer deadline makes each network response arrive earlier.

Explanation

Correct Answer

The bottleneck is outside the function CPU: examine database queries, connection pooling, caching and network architecture. Size Lambda compute to measured CPU work, not an I/O-wait floor.

Incorrect Answers

Additional CPU can improve CPU-bound phases, but a measured external latency floor does not disappear merely from larger memory allocation.

DNS zone comments are administration metadata and cannot optimize SQL execution or external service latency.

A longer timeout permits the invocation to wait longer but does not accelerate slow database calls.

Question 519

A Lambda function has reserved concurrency of 20. A burst presents 200 invocation requests, and most are throttled or held by their event source while only 20 run at once. What is the direct AWS limit causing the bottleneck?

Reserved concurrency is only a discounted billing plan and never imposes a function-level execution cap.
The function-level reserved concurrency ceiling is 20 concurrent executions; additional requests must wait, retry or be throttled by the invocation path.
Changing a queue redrive retention value automatically adds 180 reserved concurrency slots without a function configuration change.
CloudFront TTL affects the reserved concurrency limit of any unrelated Lambda function in the account.

Explanation

Correct Answer

The function cannot run 200 instances concurrently under a configured limit of 20. Event-source handling differs, but the function concurrency cap directly explains the observed limit.

Incorrect Answers

Reserved concurrency sets a hard function concurrency allocation/ceiling, independent of a price discount.

Queue retention determines message availability and retries, not the configured function concurrency ceiling.

A CDN cache policy does not change Lambda’s reserved slots for invocations.

Question 520

An API Gateway integration invokes Lambda $LATEST, while provisioned concurrency is attached only to published alias live. API traffic still shows cold starts despite billed provisioned capacity. Which mismatch explains the observation?

Provisioned concurrency automatically applies to every invocation of $LATEST even when capacity is configured for a different published alias.
Configuring function timeout to fifteen minutes silently reattaches the provisioned warm environments to $LATEST.
Reserved concurrency is identical to provisioned initialized environments, so no alias invocation qualifier is relevant.
The API is not invoking the version/alias with provisioned environments; change the integration to use the supported qualified alias.

Explanation

Correct Answer

The paid provisioned capacity lives on alias live; API Gateway traffic to $LATEST bypasses it. Invoke the correct qualified ARN and verify the intended concurrency metrics.

Incorrect Answers

Provisioned concurrency attaches to supported published versions or aliases; calls to another qualifier are not served by that capacity.

Timeout sets an execution limit and does not redirect invocation routing to the alias with preinitialized environments.

Reserved concurrency caps simultaneous execution but does not allocate preinitialized execution environments. Qualifier routing matters.

Pass Your Amazon AWS Certified Solutions Architect - Associate SAA-C03 Exam Easy!

Question 510

Explanation

Correct Answer

Incorrect Answers

Question 511

Explanation

Correct Answer

Incorrect Answers

Question 512

Explanation

Correct Answer

Incorrect Answers

Question 513

Explanation

Correct Answer

Incorrect Answers

Question 514

Explanation

Correct Answer

Incorrect Answers

Question 515

Explanation

Correct Answer

Incorrect Answers

Question 516

Explanation

Correct Answer

Incorrect Answers

Question 517

Explanation

Correct Answer

Incorrect Answers

Question 518

Explanation

Correct Answer

Incorrect Answers

Question 519

Explanation

Correct Answer

Incorrect Answers

Question 520

Explanation

Correct Answer

Incorrect Answers

AWS Certified Solutions Architect - Associate SAA-C03 Practice Tests