FinOps Cost Visibility Platform Quiz
- What is the correct order of the three FinOps operating cycle phases?
- A) Optimize → Inform → Operate
- B) Inform → Optimize → Operate
- C) Operate → Inform → Optimize
- D) Inform → Operate → Optimize
Show Answer
Answer: B) Inform → Optimize → Operate
Explanation: The FinOps cycle iterates through Inform (establish cost visibility) → Optimize (reduce costs) → Operate (governance). First understand who spends what and how much, then optimize, then manage with policies.
- What is the primary reason for integrating AWS CUR (Cost and Usage Report) with Kubecost?
- A) To reduce Kubecost license costs
- B) To track AWS service costs outside Kubernetes
- C) To improve Pod-level cost accuracy by matching actual AWS billing data
- D) To enable multi-cluster federation
Show Answer
Answer: C) To improve Pod-level cost accuracy by matching actual AWS billing data
Explanation: Kubecost estimates costs based on public list prices. CUR integration matches these estimates with actual billing data that reflects Savings Plans, Reserved Instances, and negotiated rates, significantly improving cost accuracy.
- When enforcing cost labels with Kyverno, what does
validationFailureAction: Enforcemean?- A) Show warnings for workloads without labels
- B) Block deployment of workloads without required labels
- C) Automatically add labels
- D) Modify labels on existing workloads
Show Answer
Answer: B) Block deployment of workloads without required labels
Explanation:validationFailureAction: Enforce blocks creation/modification of resources that violate the policy. Deployments without team, service, and cost-center labels will be rejected. It's recommended to start with Audit mode for warnings, then switch to Enforce when teams are ready.
- Why set VPA to
updateMode: "Off"?- A) To disable VPA entirely
- B) To provide recommendations only without automatically restarting Pods
- C) To adjust CPU only while keeping memory fixed
- D) To prevent conflicts with HPA
Show Answer
Answer: B) To provide recommendations only without automatically restarting Pods
Explanation:updateMode: "Off" makes VPA analyze resource usage and provide recommendations without automatically restarting Pods to apply changes. This supports a safe workflow where recommendations are reviewed and applied manually via PRs. The Goldilocks dashboard also leverages this mode.
- What is the difference between Showback and Chargeback?
- A) Showback displays costs, Chargeback hides costs
- B) Showback provides cost visibility, Chargeback actually charges departments/teams
- C) Showback is real-time, Chargeback is monthly
- D) Showback is cloud-only, Chargeback is on-premises only
Show Answer
Answer: B) Showback provides cost visibility, Chargeback actually charges departments/teams
Explanation: Showback shows each team/service how much they spend to raise awareness, while Chargeback actually deducts costs from department budgets. Most organizations start with Showback to establish a cost-aware culture before transitioning to Chargeback.
- What label must a namespace have for Goldilocks to display resource recommendations?
- A) goldilocks.fairwinds.com/vpa-enabled=true
- B) goldilocks.fairwinds.com/enabled=true
- C) vpa.kubernetes.io/enabled=true
- D) monitoring.goldilocks.com/watch=true
Show Answer
Answer: B) goldilocks.fairwinds.com/enabled=true
Explanation: Goldilocks automatically creates VPAs for all Deployments in namespaces labeled with goldilocks.fairwinds.com/enabled=true and visualizes recommended resource values in its web dashboard.
- What does
aggregate=label:teammean in the Kubecost Allocation API?- A) Filter only Pods with a team label
- B) Group and sum costs by team label values
- C) Create separate API calls per team
- D) Automatically add team labels
Show Answer
Answer: B) Group and sum costs by team label values
Explanation:aggregate=label:team instructs Kubecost to group all Pod costs by team label values (e.g., team-commerce, team-platform) and sum them, providing total cost, CPU cost, and memory cost per team in a single query.
- Why does the cost anomaly alert require 30 minutes of sustained high cost before firing?
- A) Prometheus scrape interval is 30 minutes
- B) To prevent false positives from temporary spikes (deployments, autoscaling)
- C) To avoid Slack API rate limits
- D) Kubecost data refresh cycle is 30 minutes
Show Answer
Answer: B) To prevent false positives from temporary spikes (deployments, autoscaling)
Explanation: Deployments, autoscaling events, and batch jobs can cause temporary cost spikes. The for: 30m condition ensures alerts fire only when costs remain elevated for 30+ minutes, reducing noise from normal operational activities.