Horizontal Pod Autoscaler in Microservices - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a Horizontal Pod Autoscaler (HPA) in Kubernetes?

HPA automatically adjusts the number of pods in a deployment based on observed CPU utilization or other select metrics to maintain application performance.

intermediate

Which metrics can Horizontal Pod Autoscaler use to scale pods?

HPA can use CPU utilization, memory usage, or custom metrics like request rate to decide when to scale pods up or down.

intermediate

How does HPA decide when to increase or decrease pod count?

HPA compares current metric values against target thresholds. If usage is above the target, it adds pods; if below, it removes pods to optimize resource use.

advanced

What is the difference between Horizontal Pod Autoscaler and Vertical Pod Autoscaler?

Horizontal Pod Autoscaler changes the number of pods, while Vertical Pod Autoscaler changes the resource requests and limits (CPU/memory) of existing pods.

intermediate

Why is it important to set minimum and maximum pod limits in HPA?

Setting min and max pod limits prevents scaling too low (causing performance issues) or too high (wasting resources), ensuring stable and efficient operation.

What does Horizontal Pod Autoscaler primarily adjust in a Kubernetes cluster?

A Network bandwidth

B CPU limits of pods

C Memory limits of pods

D Number of pods

HPA adjusts the number of pods to handle workload changes.

Which metric is commonly used by HPA to trigger scaling?

A Disk space usage

B CPU utilization

C Number of nodes

D Pod restart count

CPU utilization is a common metric for HPA to decide scaling.

What happens if the current CPU usage is below the target in HPA?

A Pods are removed

B Pods remain the same

C Pods are added

D Nodes are added

If usage is below target, HPA removes pods to save resources.

Which Kubernetes component typically manages the Horizontal Pod Autoscaler?

A kube-controller-manager

B kube-scheduler

C kube-proxy

D etcd

The kube-controller-manager runs the HPA controller.

Why should you avoid setting the maximum pod count too high in HPA?

A It slows down pod startup

B It can cause pod starvation

C It wastes cluster resources

D It causes network congestion

Too high max pods can waste resources and increase costs.

Explain how Horizontal Pod Autoscaler works to maintain application performance.

Think about how a thermostat adjusts temperature automatically.

Describe the key differences between Horizontal Pod Autoscaler and Vertical Pod Autoscaler.

One changes quantity, the other changes size.