CPU requests and limits in Kubernetes - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding CPU Requests vs Limits

In Kubernetes, what happens when a container exceeds its CPU request but stays below its CPU limit?

A The container is allowed to use unlimited CPU resources.

B The container is throttled to stay within the CPU limit but can use more CPU than requested.

C The container is paused until CPU usage drops below the request.

D The container is immediately terminated for exceeding the CPU request.

💡 Hint

Think about how Kubernetes manages guaranteed CPU resources versus maximum allowed CPU.

CPU requests define the guaranteed CPU for a container. If the container uses more CPU than requested but less than the limit, Kubernetes allows it but may throttle it to not exceed the limit.

💻 Command Output

intermediate

CPU Limit Enforcement Output

Given this pod spec snippet with CPU requests and limits, what will be the CPU usage behavior if the container tries to use 2 CPUs but the limit is set to 1 CPU?

resources:
  requests:
    cpu: "500m"
  limits:
    cpu: "1"

A The container will be throttled and cannot use more than 1 CPU.

B The container will use 2 CPUs without restriction.

C The container will be killed for exceeding the CPU limit.

D The container will only use 500m CPU and no more.

💡 Hint

CPU limits set the maximum CPU a container can use.

The CPU limit of 1 CPU means the container cannot use more than 1 CPU. If it tries to use 2 CPUs, Kubernetes throttles it to 1 CPU.

❓ Configuration

advanced

Correct CPU Resource Configuration

Which of the following Kubernetes container resource configurations correctly sets a CPU request of 250m and a CPU limit of 500m?

A

resources:
  requests:
    cpu: "250m"
  limits:
    cpu: "500m"

B

resources:
  requests:
    cpu: 250
  limits:
    cpu: 500

C

resources:
  requests:
    cpu: "0.25"
  limits:
    cpu: "0.5"

D

resources:
  requests:
    cpu: "250"
  limits:
    cpu: "500"

💡 Hint

CPU values in Kubernetes are usually specified in millicores with 'm' suffix.

CPU requests and limits should be specified as strings with units. "250m" means 250 millicores (0.25 CPU). Options A and D lack units and are invalid. Option A uses decimal notation which is valid but less common; however, Kubernetes expects strings for resource quantities.

❓ Troubleshoot

advanced

Diagnosing CPU Throttling Issues

A developer complains their pod is running slower than expected. The pod has CPU requests and limits set. Which kubectl command helps check if the pod is being CPU throttled?

A kubectl exec <pod-name> -- top

B kubectl get pods --all-namespaces

C kubectl logs <pod-name>

D kubectl describe pod <pod-name>

💡 Hint

Look for events and resource usage details in pod descriptions.

The kubectl describe pod command shows detailed pod info including events and resource usage, which can indicate CPU throttling. Other commands do not provide this info directly.

🔀 Workflow

expert

Optimizing CPU Requests and Limits for a Production Pod

You want to optimize CPU resource settings for a production pod that sometimes spikes in CPU usage but should not be throttled during spikes. Which approach is best?

A Set CPU request to maximum spike and no CPU limit.

B Set CPU request and limit both to the maximum expected spike.

C Set CPU request to average usage and CPU limit to maximum expected spike.

D Set CPU request to zero and CPU limit to maximum spike.

💡 Hint

Think about balancing guaranteed resources and allowing burst capacity.

Setting CPU request to average usage guarantees baseline CPU. Setting limit to max spike allows bursting without throttling. Setting request and limit equal wastes resources. No limit risks node stability. Zero request risks eviction.