Find your next open-source project

Search by name, purpose, or keyword — or paste a GitHub repository URL
All skills

gke-workload-scaling

Configure HPA and VPA to scale GKE workloads and right-size pod resources.

  • Part of google/skills
  • 6.5k installs
  • DevOps & Self-hosting · Containers and clusters

When to use it

You have a GKE deployment with resource requests set but no autoscaling policy, and need to add HPA or VPA.

For example

You say

Use gke-workload-scaling to add an HPA to my checkout-api deployment in the prod namespace, scaling on CPU between 2 and 10 pods.

You get
  • HPA YAML manifest
  • kubectl apply command
  • HPA verification steps

How it works

  • VPA Off mode runs 24+ hours to collect recommendations before applying a 20% buffer.
  • HPA and VPA should not share the same metric; a common split is HPA on CPU and VPA on memory.

Before installing

VPA on a Standard cluster must be enabled first with a gcloud command; Autopilot clusters have it by default.