Kubernetes Cost Optimization

Optimize Kubernetes costs through resource right-sizing, unused resource detection, and cluster efficiency analysis. Use for cost optimization, resource analysis, and capacity planning.

Cost analysis and optimization using kubectl-mcp-server's cost tools.

Requires the cost, keda toolsets: start kubectl-mcp-server with --toolsets default,cost,keda (or MCP_TOOLSETS=default,cost,keda).

When to Apply

Use this skill when:

• User mentions: "cost", "savings", "optimize", "expensive", "budget"
• Operations: cost analysis, right-sizing, cleanup unused resources
• Keywords: "how much", "reduce", "efficiency", "waste", "overprovisioned"

Priority Rules

Priority	Rule	Impact	Tools
1	Find and delete unused PVCs	CRITICAL	get_idle_resources
2	Right-size overprovisioned pods	HIGH	get_resource_recommendations
3	Identify idle LoadBalancers	HIGH	kubectl_get
4	Scale down non-prod off-hours	MEDIUM	kubectl_scale
5	Consolidate small namespaces	LOW	Analysis

Quick Reference

Task	Tool	Example
Namespace cost	get_namespace_cost_allocation	get_namespace_cost_allocation(namespace)
Cluster cost	get_cost_analysis	get_cost_analysis()
Unused PVCs	get_idle_resources	get_idle_resources(namespace)
Right-sizing	get_resource_recommendations	get_resource_recommendations(namespace)

Quick Cost Analysis

Get Cost Summary

python

get_namespace_cost_allocation(namespace)
get_cost_analysis()

Find Unused Resources

python

get_idle_resources(namespace)
get_idle_resources(namespace)

Resource Right-Sizing

python

get_resource_recommendations(namespace)
kubectl_top(name, namespace)

Cost Optimization Workflow

1. Identify Overprovisioned Resources

python

get_resource_recommendations(namespace="production")

kubectl_top(name, namespace)
kubectl_top(namespace)

2. Find Idle Resources

python

get_idle_resources(namespace)

get_idle_resources(namespace)

3. Analyze Node Utilization

python

kubectl_get(resource_type="nodes")
kubectl_top(resource="nodes")

Right-Sizing Guidelines

Current State	Recommendation
CPU usage < 10% of request	Reduce request by 50%
CPU usage > 80% of request	Increase request by 25%
Memory < 50% of request	Reduce request
Memory near limit	Increase limit, monitor OOM

Cost by Resource Type

Compute (Pods/Deployments)

python

kubectl_top(namespace)
kubectl_top(name, namespace)

Storage (PVCs)

python

kubectl_get(resource_type="pvc", namespace)
get_idle_resources(namespace)

Network (LoadBalancers)

python

kubectl_get(resource_type="services", namespace)

Multi-Cluster Cost Analysis

Compare costs across clusters:

python

get_cost_analysis(context="production")
get_cost_analysis(context="staging")
get_cost_analysis(context="development")

Cost Reduction Actions

Immediate Wins

1. Delete unused PVCs: get_idle_resources() then delete
2. Right-size pods: Apply get_resource_recommendations()
3. Scale down dev/staging: Off-hours scaling

Medium-term Optimizations

1. Use Spot/Preemptible nodes: For fault-tolerant workloads
2. Implement HPA: Auto-scale based on demand
3. Use KEDA: Scale to zero for event-driven workloads

Long-term Strategy

1. Reserved instances: For stable production workloads
2. Multi-tenant clusters: Consolidate small clusters
3. Right-size node pools: Match workload requirements

Automated Analysis Script

For comprehensive cost analysis, see scripts/find-overprovisioned.py.

KEDA for Cost Savings

Scale to zero with KEDA:

python

keda_scaledobjects_list(namespace)
keda_scaledobject_get(name, namespace)

KEDA reduces costs by:

• Scaling pods to 0 when idle
• Event-driven scaling (queue depth, etc.)
• Cron-based scaling for predictable patterns

Related Skills

• k8s-autoscaling - HPA, VPA, KEDA
• k8s-troubleshoot - Resource debugging