Capacity Planning
Analyze actual infrastructure utilization patterns and model infrastructure changes before committing. Compare instance types, simulate Graviton migration, optimize spot ratios, and identify peak waste windows with instant cost projections for every scenario.
How it Works
Three steps from setup to savings
Baseline
Historical utilization data (up to 360 days) is ingested and normalized across all clusters and cloud providers.
Analyze
Utilization patterns are analyzed by node group, instance type, zone, and hour of day to surface peak waste windows and efficiency gaps.
Model
Use the what-if simulator to evaluate Graviton migration, spot instance ratios, and node pool consolidation scenarios with instant cost projections.
What's Included
Infrastructure Utilization Analysis & Scenario Modeling
Utilization Pattern Analysis
Analyze infrastructure utilization across node groups, instance types, zones, and time of day to find where capacity is idle or wasted.
- Hourly utilization heatmaps reveal peak and off-peak capacity patterns across your cluster
- Spot vs on-demand mix and cost-per-instance-type breakdown across all node groups
Utilization Thresholds
Set target utilization bands and get alerted when actual utilization approaches capacity boundaries.
- Define target utilization bands (e.g., 60–80%) for CPU and memory per node group
- Proactive alerts when current utilization trends toward your configured capacity limits
Node Pool Planning
Model different instance types and quantities to find the most cost-effective capacity configuration.
- Compare instance types by cost-per-vCPU and cost-per-GB to find optimal mix
- Model node pool configurations with spot, on-demand, and reserved blends
What-If Scenario Simulator
Model three infrastructure scenarios with instant cost projections: Graviton/ARM migration, spot instance ratio adjustment, and node pool consolidation.
- Graviton migration (~20% savings), spot ratio tuning (~60% savings), consolidation (~10% savings) compared side-by-side
- Current vs projected monthly cost shown instantly as you adjust scenario parameters
Multi-Cluster View
Aggregate capacity metrics across clusters for a unified view of infrastructure headroom.
- Unified capacity dashboard across all clusters regardless of cloud provider
- Identify which clusters have headroom and which need immediate scaling
Scaling Recommendations
Get specific recommendations for HPA targets, node pool sizes, and autoscaler configurations.
- Specific HPA min/max replica recommendations based on traffic analysis
- Node autoscaler config suggestions for optimal bin-packing efficiency
Frequently Asked Questions
Common questions about Capacity Planning
Ready to Start Your
Kubernetes FinOps Journey?
Stop overpaying for Kubernetes. See potential savings within 10 minutes.