KCAVO - Kubernetes Cluster Analyse Visualize and Optimize

A kubectl or Kubernetes Command Line Tool based plugin that visualizes resources, analyzes costs, and provides optimization recommendations.

What it helps you with?

• Cost Analysis: Calculate and visualize costs for pods, nodes, and resources
• GPU Analysis: Specialized GPU resource tracking and optimization
• Resource Visualization: View cluster resources in clean tables
• Smart Recommendations: Get actionable cost-saving suggestions

Install

make install
kubectl cost --version

Quick Start

# Analyze costs in current namespace
kubectl cost analyze

# Analyze across all namespaces
kubectl cost analyze --all-namespaces

# Visualize cluster resources
kubectl cost visualize

# Analyze GPU usage
kubectl cost gpu

# Get optimization recommendations
kubectl cost optimize

Commands

kubectl cost analyze

Analyze costs for pods and resources in your cluster.

# Basic usage
kubectl cost analyze

# Across all namespaces
kubectl cost analyze -A

# With proper breakdown
kubectl cost analyze --breakdown

# Show top 10 expensive
kubectl cost analyze --top 10 --sort-by cost

# Different output formats
kubectl cost analyze -o json
kubectl cost analyze -o yaml

kubectl cost visualize

Visualize Kubernetes resources in table format.

# Visualize all resources
kubectl cost visualize

# Specific resource type
kubectl cost visualize --type pods
kubectl cost visualize --type nodes

# All namespaces
kubectl cost visualize -A

kubectl cost gpu

Analyze GPU resource allocation and usage.

# Analyze GPU usage
kubectl cost gpu

# All namespaces
kubectl cost gpu -A

kubectl cost optimize

Get cost optimization recommendations.

# Get recommendations
kubectl cost optimize

# All namespaces
kubectl cost optimize -A

Configuration

Create ~/.kubectl-cost.yaml to customize pricing or use existing cloud providers' pricing:

pricing:
  cpu_hourly: 0.024           # $17.28/month per core
  memory_gb_hourly: 0.003     # $2.16/month per GB
  gpu_hourly: 0.90            # $648/month per GPU
  storage_gb_monthly: 0.10    # $0.10/month per GB

# Or use cloud provider presets
provider: aws  # aws, gcp, or azure

Calculation for Costs

Costs are calculated based on:

• CPU: Resource requests (or limits if requests not set)
• Memory: Resource requests (or limits if requests not set)
• GPU: GPU resource requests
• Time: Monthly basis (730 hours/month)

Formula:

Monthly Cost = (CPU_cores × CPU_rate + Memory_GB × Memory_rate + GPU_count × GPU_rate) × 730_hours

Acknowledgements

• https://github.com/spf13/cobra for CLI.
• https://github.com/olekukonko/tablewriter for the tables.