Scaling Strategies Quiz
Related Document: Scaling Strategies
Multiple Choice Questions
1. What is required to use custom metrics with HPA?
- A) Kubernetes 1.30 or higher
- B) Prometheus Adapter or similar custom metrics server
- C) AWS Auto Scaling integration
- D) Manual pod scaling
Show Answer
Answer: B) Prometheus Adapter or similar custom metrics server
Explanation: HPA with custom metrics requires a metrics server that implements the custom.metrics.k8s.io API. Prometheus Adapter queries Prometheus and exposes metrics in the format HPA expects, enabling scaling on application-specific metrics like requests per second.
2. What type of events can KEDA scale on that HPA cannot?
- A) CPU utilization
- B) Memory usage
- C) External events like SQS queue depth or Kafka lag
- D) Pod restart counts
Show Answer
Answer: C) External events like SQS queue depth or Kafka lag
Explanation: KEDA (Kubernetes Event-Driven Autoscaling) includes scalers for external systems like AWS SQS, Kafka, RabbitMQ, databases, and more. It can scale to zero and react to events outside the Kubernetes metrics system.
3. What is the difference between VPA modes: "Auto" and "Off"?
- A) Auto enables HPA, Off disables it
- B) Auto updates pods in-place, Off only provides recommendations
- C) Auto requires restart, Off is instant
- D) Auto uses Spot instances, Off uses On-Demand
Show Answer
Answer: B) Auto updates pods in-place, Off only provides recommendations
Explanation: VPA "Auto" mode automatically evicts and recreates pods with updated resource requests. "Off" mode only generates recommendations without making changes, useful for reviewing suggestions before manual implementation.
4. What is Pod Deletion Cost and how does it affect scaling?
- A) The financial cost of deleting a pod
- B) An annotation that influences which pods are removed first during scale-down
- C) The time required to delete a pod
- D) Storage costs associated with pod termination
Show Answer
Answer: B) An annotation that influences which pods are removed first during scale-down
Explanation: The controller.kubernetes.io/pod-deletion-cost annotation assigns a cost value to pods. During scale-down, pods with lower deletion cost are terminated first. This helps preserve pods running important workloads or those with cached data.
5. When using HPA with custom metrics, what is the formula for calculating desired replicas?
- A) currentReplicas + 1
- B) desiredReplicas = ceil(currentReplicas * (currentMetricValue / desiredMetricValue))
- C) maxReplicas / 2
- D) currentMetricValue * targetUtilization
Show Answer
Answer: B) desiredReplicas = ceil(currentReplicas * (currentMetricValue / desiredMetricValue))
Explanation: HPA calculates desired replicas by comparing current metric values to target values, then scaling proportionally. The ceiling function ensures at least one additional replica is added when scaling up.
6. What is the recommended strategy for Spot node interruption handling?
- A) Ignore interruptions
- B) Use Pod Disruption Budgets and graceful termination with interruption handlers
- C) Only run stateless workloads
- D) Disable Spot instances entirely
Show Answer
Answer: B) Use Pod Disruption Budgets and graceful termination with interruption handlers
Explanation: Spot interruption handling combines AWS Node Termination Handler (or Karpenter's native handling), Pod Disruption Budgets to ensure availability, adequate termination grace periods, and workload design that handles preemption gracefully.
7. In KEDA, what does pollingInterval configure?
- A) How often pods are restarted
- B) How frequently KEDA checks the external metric source
- C) The delay between scale operations
- D) Health check frequency
Show Answer
Answer: B) How frequently KEDA checks the external metric source
Explanation:pollingInterval defines how often KEDA queries the external scaler (e.g., SQS, Prometheus) for metric values. Lower intervals provide faster reaction times but increase load on the metric source.
8. What Kubernetes feature enables VPA to resize pods without restart (in newer versions)?
- A) Rolling updates
- B) In-place Pod Vertical Scaling
- C) Blue/green deployment
- D) Canary releases
Show Answer
Answer: B) In-place Pod Vertical Scaling
Explanation: Kubernetes 1.27+ supports in-place vertical scaling through the resizePolicy field, allowing CPU and memory changes without pod restart. VPA can leverage this for less disruptive resource adjustments.
9. What is the benefit of using Prometheus Adapter over metrics-server for HPA?
- A) Lower resource usage
- B) Support for custom and external metrics beyond CPU/memory
- C) Faster metric collection
- D) Built-in alerting
Show Answer
Answer: B) Support for custom and external metrics beyond CPU/memory
Explanation: Metrics-server only provides CPU and memory metrics. Prometheus Adapter exposes any Prometheus metric through the custom.metrics.k8s.io API, enabling HPA to scale on application metrics like requests per second, queue depth, or latency.
10. What should you consider when combining HPA and VPA?
- A) They cannot be used together
- B) Configure them on different resource dimensions (HPA on CPU, VPA on memory)
- C) Always disable HPA when VPA is enabled
- D) They automatically coordinate without configuration
Show Answer
Answer: B) Configure them on different resource dimensions (HPA on CPU, VPA on memory)
Explanation: HPA and VPA can conflict if both try to manage the same resource dimension. A common pattern is using HPA for horizontal scaling based on CPU and VPA in recommendation mode for memory, or using the VPA "Initial" mode which only sets resources at pod creation.