mcpField
G
gke-cost-optimization

gke-cost-optimization

github.com/googlecloudplatformCloud & DevOps16947 installsUpdated 3d ago
Repository

About gke-cost-optimization

gke-cost-optimization is a cost guidance skill for people operating GKE clusters. It covers cost allocation, resource quotas, and workload rightsizing to help track and control resource spending.

Key features

  1. 01Enable GKE cost allocation
  2. 02View costs by namespace or label
  3. 03Configure resource quotas
  4. 04Use VPA to recommend resource requests
  5. 05Coordinate HPA and VPA for workloads

Use cases

  1. 01Analyze cluster and namespace costs
  2. 02Limit resource use in multi-tenant clusters
  3. 03Adjust workloads to observed utilization
  4. 04Use Spot VMs for fault-tolerant stateless workloads

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/googlecloudplatform/gke-mcp --skill gke-cost-optimization

After installing, type/gke-cost-optimizationin Claude Code to confirm it loaded.

Details

From the project README

GKE Cost Optimization

This skill provides guidance on optimizing costs for Google Kubernetes Engine (GKE) clusters.

Overview

Cost optimization in GKE involves tracking costs, setting limits to prevent waste, and rightsizing workloads to match actual usage.

Workflows

1. Enable GKE Cost Allocation

GKE cost allocation allows you to see the cost of your GKE resources in Cloud Billing, broken down by namespace and cluster labels.

Steps:

  1. Enable GKE cost allocation in the cluster settings.

Command:

gcloud container clusters update <cluster-name> \
    --enable-cost-allocation \
    --region <region>
  1. View costs in the Cloud Billing reports by grouping by namespace or labels.
2. Configure Resource Quotas

Resource quotas restrict the total resource consumption in a namespace, preventing any single tenant from consuming all cluster resources.

Example ResourceQuota Manifest:

apiVersion: v1
kind: ResourceQuota
metadata:
  name: compute-quota
  namespace: my-namespace
spec:
  hard:
    requests.cpu: "4"
    requests.memory: 16Gi
    limits.cpu: "8"
    limits.memory: 32Gi
3. Rightsizing Strategies

Rightsizing involves adjusting the requested resources of your workloads to match their actual utilization.

  • Use VPA in Recommender Mode: Let VPA observe usage and recommend CPU and memory requests.
  • Use MPA: Reconcile HPA and VPA to avoid conflicts.
  • Review Cost Recommendations: Check the Google Cloud Console for GKE cost optimization recommendations.

Best Practices

  1. Enable Cost Allocation: Always enable GKE cost allocation to understand where your money is going.
  2. Use Resource Quotas: Enforce resource quotas in multi-tenant clusters to prevent cost runaways.
  3. Leverage Spot VMs: Use Spot VMs for fault-tolerant, stateless workloads to save up to 91%.
  4. Automate Scaling: Use Cluster Autoscaler and HPA/VPA to ensure you only pay for what you need.

Comments

Markdown supported