Question 160
Product costs span multiple linked accounts and shared platform resources. The finance team needs to report expenses using business product categories that combine multiple cost allocation tags and account groups. Which billing tool most directly builds those categories?
- Create individual AWS Budgets scoped to each payer-linked account and use their alert amounts as the multi-dimensional product ledger.
- Use AWS Cost Categories with rules combining relevant account, service and activated cost allocation tag dimensions.
- Activate one product cost-allocation tag and report only those tagged resources, excluding charges from shared accounts that need business allocation.
- Use Cost Explorer grouping by service and linked account each month, but rebuild product ownership manually in spreadsheets instead of defining reusable rules.
Explanation
Correct Answer
B:Cost Categories create business-oriented groupings across billing dimensions for allocation and analysis while preserving centralized reporting.
Incorrect Answers
A:Budgets can detect and forecast spend, but they do not implement the requested rules that combine accounts, services, tags and shared-platform categories.
C:A single tag can track direct spend but does not supply the composite classification and allocation rules for shared product costs described in the question.
D:Grouping is useful for ad hoc analysis, yet it does not centrally codify the repeatable business product categories requested across shared resources and accounts.
Question 161
A data team wants to query granular line items of AWS usage and cost using SQL, including resource identifiers where supported, rather than manually export a monthly console chart. What is the most direct reporting architecture?
- Use Cost Explorer console reports exported to CSV and query them as if they contain the complete supported resource-level CUR line-item schema.
- Deliver an AWS Cost and Usage Report or equivalent detailed AWS billing export to S3 and query the dataset with Athena.
- Export AWS Config aggregator snapshots to S3 and compute billing charges from each configuration change timestamp.
- Send organization CloudTrail management events to Athena and calculate every AWS charge from its API calls alone.
Explanation
Correct Answer
B:A detailed export in S3 provides line-item usage and cost dimensions for SQL analysis; Athena can query a suitably cataloged report.
Incorrect Answers
A:Cost Explorer summaries differ from granular line-item exports and do not necessarily contain the requested full resource identifiers and usage dimensions.
C:Config records resource state, not the authoritative billed prices, savings plan allocations and measured usage line items of CUR.
D:CloudTrail records API activity but not complete metered usage, recurring fees, credit application and discounted costs across services.
Question 162
An enterprise runs EC2, Fargate and Lambda usage with a stable compute-spend baseline but shifting instance families and Regions. It seeks a flexible spend commitment rather than a family-specific capacity reservation. Which choice best fits?
- Buy an EC2 Instance Savings Plan restricted to one instance family and Region, then treat eligible Fargate and Lambda usage as covered by the commitment.
- Purchase Standard Reserved Instances for the dominant EC2 family and expect those reservations to automatically discount all serverless usage.
- Evaluate a Compute Savings Plan against the stable eligible hourly spend and use coverage/utilization reports to size its commitment.
- Purchase a much larger Compute Savings Plan without measuring the stable eligible spend or existing commitment coverage.
Explanation
Correct Answer
C:Compute Savings Plans can cover eligible EC2, Fargate and Lambda usage while allowing more compute flexibility than narrowly scoped instance commitments.
Incorrect Answers
A:EC2 Instance Savings Plans apply to a particular instance family and Region, not to Fargate and Lambda; their tighter scope does not meet the workload flexibility requirement.
B:EC2 reservations can suit stable EC2 demand but do not offer the cross-family, cross-Region and eligible Fargate/Lambda flexibility of a Compute Savings Plan.
D:The plan type is suitable but an oversized dollar-per-hour commitment can create unused spend. Historical coverage and utilization are required to size the commitment responsibly.
Question 163
A support organization has many idle nonproduction EC2 instances that teams forget to stop on weekends. Spending reports show the waste occurs during known off-hours and no weekend testing is authorized. What targeted control should the architect recommend?
- Apply validated schedule-based stop/start automation to tagged nonproduction instances, with opt-outs and stateful-workload safeguards.
- Buy a new Compute Savings Plan to discount all nonproduction instance-hours while continuing to run the fleet through unused weekends.
- Use EC2 Auto Scaling target tracking alone for these manually provisioned instances, assuming all will stop at midnight regardless of demand.
- Apply an AWS Budgets action that limits future instance launches when the monthly forecast rises, but do not stop existing idle instances.
Explanation
Correct Answer
A:Scheduling targets the measured idle runtime hours directly while preserving exceptions for workloads that cannot safely stop and restart.
Incorrect Answers
B:A commitment can reduce unit pricing but does not eliminate idle runtime; scheduled stops target the observed avoidable consumption directly.
C:Auto Scaling governs supported group capacity from metrics, not an explicit weekend stop schedule for unrelated existing instances and their state safeguards.
D:A launch restriction can limit future spending but fails to address the already-running instances that account for the known weekend waste.
Question 164
A company bought a one-year commitment sized against last year’s EC2 utilization. Current Cost Explorer reports show the commitment is often underutilized after workload rightsizing. What should finance and architecture do before buying additional discounts?
- Purchase additional Compute Savings Plans because coverage is low, ignoring the cost of the existing unused hourly commitment.
- Compare only instance-family On-Demand rates after rightsizing, rather than reviewing effective discounted spend and utilization.
- Analyze Savings Plans utilization and coverage against current usage forecasts, then reduce or defer new commitments accordingly.
- Convert the underused commitment into a larger EC2 Instance Savings Plan without first validating a stable family-specific workload.
Explanation
Correct Answer
C:Utilization and coverage reveal unused commitment and remaining eligible on-demand spend; recent rightsizing changes the spend baseline.
Incorrect Answers
A:Low coverage alone does not mean another commitment is worthwhile; existing low utilization after rightsizing may mean the organization already overcommitted.
B:List-price comparisons omit the existing hourly commitments, which determine whether additional plans create net savings or further unused charges.
D:Tighter family-specific coverage can magnify underutilization and does not follow from evidence that the previous purchase already exceeds current demand.
Question 165
Finance sees rising cloud spend but cannot determine which recently deployed resource group accounts for the change. Production accounts have owner and cost-center tags, but the billing export does not include those tag dimensions. What should the organization fix?
- Enable AWS Cost Explorer resource-level cost views without activating the existing user-defined owner and cost-center tags in Billing.
- Apply an AWS Cost Category that groups accounts by service while leaving the owner and cost-center tag dimensions unactivated.
- Activate the relevant user-defined cost allocation tags and verify that governed tagging appears in subsequent billing datasets.
- Turn on CloudTrail resource event logging and infer the cost center from the last principal that modified each resource.
Explanation
Correct Answer
C:Resource tags are not necessarily available as cost-allocation dimensions until activated in billing, after which analysis can use the tagged expenses.
Incorrect Answers
A:Cost Explorer can display cost allocations only for applicable dimensions; user-defined tag values need activation to appear in the billing datasets.
B:Account/service grouping can assist finance but cannot expose the missing tag values in CUR until the relevant cost allocation tags are activated.
D:The last API caller does not reliably identify chargeback ownership and is not a replacement for activated resource cost allocation tags.
Question 166
An e-commerce group has an unexpected overnight jump in AWS costs caused by a newly created service in a member account. Finance wants prompt alerts on anomalous spending before the next monthly invoice rather than only static monthly caps. Which cost tool fits?
- Configure only a fixed AWS Budgets monthly alert threshold for the organization and assume it detects any unusual spending pattern early.
- Configure AWS Cost Anomaly Detection with appropriate monitors and alert subscriptions for the account or service scope.
- Use Cost Explorer to inspect yesterday’s spending manually without enabling anomaly monitor subscriptions.
- Configure an AWS Cost Category for the new service and assume category assignment itself creates anomaly alerts.
Explanation
Correct Answer
B:Anomaly Detection observes cost patterns and notifies stakeholders about unusual spend changes that static budget estimates might miss.
Incorrect Answers
A:A static budget can flag a threshold but may miss smaller anomalous changes in a member service before that dollar limit is reached.
C:Historical cost reports require manual checking; they are not the automated unusual-spend detection requested.
D:Cost Categories classify charges but do not detect unexpected changes without a separate anomaly monitor/alert subscription.
Question 167
An enterprise has steady eligible compute spending but deployment teams frequently change EC2 instance families and use Lambda and Fargate as needed. A finance manager wants commitment discounts without choosing one EC2 family for every workload. What purchasing approach is best to evaluate?
- Purchase EC2 Instance Savings Plans scoped to one selected family and Region for all usage, including unrelated Lambda and Fargate compute.
- Purchase an On-Demand Capacity Reservation for a chosen EC2 instance size and assume it discounts all serverless compute.
- Analyze an appropriately sized Compute Savings Plan and track its hourly utilization and coverage against eligible workloads.
- Buy an EC2 Instance Savings Plan for one narrowly selected family and assume it covers all Lambda invocations.
Explanation
Correct Answer
C:Compute Savings Plans offer broad eligible compute coverage across EC2, Fargate and Lambda, with spend rather than a single instance family as the commitment basis.
Incorrect Answers
A:EC2 Instance Savings Plans can discount a specific EC2 family and Region but do not cover general Lambda/Fargate usage or shifts among families.
B:Capacity Reservations reserve matching EC2 capacity; they are not flexible Compute Savings Plans across EC2, Lambda and Fargate.
D:An EC2 Instance Savings Plan has narrower coverage than a Compute Savings Plan and cannot cover the mixed workloads as stated.
Question 168
A central EKS cluster runs workloads for several product teams on shared EC2 worker nodes. The ordinary AWS Cost and Usage Report attributes node charges to the same EC2 instances, but Finance needs pod and namespace-level CPU/memory cost allocation for internal chargeback without counting the same node cost twice. Which reporting configuration should be enabled?
- Use only the parent EC2 instance’s Cost Explorer total and assign its full price to every namespace, which intentionally duplicates shared costs.
- Opt into AWS split cost allocation data for Amazon EKS in the billing export and analyze generated per-pod resource allocation and tags.
- Buy a dedicated Reserved Instance for each Kubernetes pod, even when containers share the same EC2 host and do not map to EC2 instance IDs.
- Export only the EKS cluster name as a resource tag and assume the resulting EC2 line item identifies each pod’s actual CPU and memory consumption.
Explanation
Correct Answer
B:Split cost allocation data adds pod-level CPU and memory attribution to CUR so shared EKS-on-EC2 worker spend can be charged back.
Incorrect Answers
A:Duplicating one node charge for every namespace overstates cost instead of allocating shared capacity among actual workloads.
C:Reserved Instances attach to eligible compute usage rather than individual Kubernetes pods sharing a single EC2 instance.
D:A cluster tag can label the host allocation but cannot split shared node cost according to each pod’s measured resource usage.
Question 169
An AWS Cost Category already classifies direct cloud spend for Product A and Product B and places shared CI/CD and DNS expense in a Shared-Platform category. Finance must allocate the Shared-Platform source costs 70/30 to Product A/B for an internal chargeback report. It accepts that these allocations appear in the Cost Categories details view rather than being rewritten into CUR. Which feature supplies that distribution?
- Cost Category split charge rules with Shared-Platform as the source and Product A/B as targets using the agreed allocation proportion.
- Define two more Cost Categories with the same Shared-Platform classification rules and treat them as a 70/30 expense split.
- Use an AWS Budgets forecast with a cost-center filter to reassign 70 percent of each invoice line to Product A.
- Enable amortized-cost reporting in Cost Explorer and assume the amortization setting alone splits platform expense by product ratio.
Explanation
Correct Answer
A:Split charge rules apportion a shared cost category value across target values, with supported fixed/percentage/proportional methods in the category details.
Incorrect Answers
B:Additional categories merely classify costs; without actual split charge rules the shared source value is not allocated by the required ratio.
C:AWS Budgets tracks and alerts on spending but does not rewrite or allocate shared Cost Category source values.
D:Amortization changes the time treatment of commitment charges but does not establish business allocation ratios for shared services.
Question 170
Finance has one compute Savings Plan with 99% commitment utilization, but the coverage report shows that only 35% of eligible predictable EC2, Fargate and Lambda usage receives a Savings Plans discount. Historical eligible spending has remained stable and workload forecasts show no imminent rightsizing. What conclusion should drive the next commitment calculation?
- The plan has low utilization because a 35% coverage rate is mathematically identical to commitment utilization, so cancel all new commitments.
- Evaluate additional commitment against the uncovered stable eligible hourly spend and forecast; high utilization and low coverage can coexist.
- Purchase a Reserved DB Instance instead because it applies its discount automatically to all EC2, Fargate and Lambda compute usage.
- Use a Cost Category allocation rule as a billing discount purchase mechanism that replaces a Savings Plan commitment.
Explanation
Correct Answer
B:A fully used existing commitment with substantial uncovered predictable eligible usage may justify carefully sized incremental commitments after forecast review.
Incorrect Answers
A:Coverage measures the share of eligible usage discounted, whereas utilization measures how fully a purchased hourly commitment is consumed.
C:Reserved DB Instances are an RDS pricing instrument, not a substitute for flexible EC2, Fargate and Lambda Savings Plans discounts.
D:Cost Categories attribute billing data to business dimensions; they do not create a new discounted compute commitment.
AWS Certified Solutions Architect - Professional SAP-C02 Practice Tests