/scaffold

Terraform Cost Estimation Reference

Purpose: Use this file when you need a cost estimate, warning thresholds, tagging rules, or budget patterns for Terraform-managed infrastructure.

Contents:

1. Estimation workflow
2. Infracost quick start
3. High-cost signals
4. Environment multipliers and formulas
5. Report template
6. Budgets, tagging, commitments, optimization

Estimation Workflow

1. Read Terraform files and identify billable resource blocks.
2. Extract cost-driving attributes: size, count, AZ count, HA flags, transfer, storage, min_instances.
3. Apply environment multipliers:
   • dev: single-AZ, smaller, no HA when safe
   • staging: production-like but scaled down
   • prod: full HA, backups, monitoring
4. Produce:
   • per-resource breakdown
   • category subtotal
   • warning list

Infracost Quick Start

brew install infracost
infracost auth login
infracost breakdown --path /path/to/terraform
infracost diff --path /path/to/terraform --compare-to infracost-base.json

Monorepo example:

version: 0.1
projects:
  - path: environments/dev
    name: dev
  - path: environments/staging
    name: staging
  - path: environments/prod
    name: prod

Note: the free tier referenced in the existing skill is 1,000 runs/month.

High-Cost Signals

Flag these explicitly:

Environment Multipliers And Formulas

Key heuristics:

• AWS Graviton / GCP Tau can reduce compute cost by about 20%
• Cloud SQL REGIONAL is about 1.8x ZONAL
• Savings Plans / RI / CUD can reduce steady-state cost by roughly 28-72% depending on commitment type
• Spot / Preemptible can reduce cost by about 60-90% for interruptible workloads

Useful formulas:

Fargate monthly
= desired_count × (vCPU_price × vCPU + memory_price × memory_GB) × 730

Lambda monthly
= request_cost + GB-seconds × compute_rate

Aurora Serverless v2 monthly
= avg_ACU × ACU_rate × 730 + storage_GB × storage_rate

Cloud Run monthly
= active_seconds × per-second price + request charges

Cost Report Template

## Cost Estimate: [Project/Module Name]

**Provider**: AWS / GCP / Azure
**Region**: [region]
**Estimated by**: Manual analysis / Infracost
**Date**: YYYY-MM-DD

### Resource Breakdown

| # | Resource | Terraform Reference | Spec | Count | Monthly (USD) |
|---|----------|-------------------|------|-------|---------------|
| 1 | [type]   | `resource.name`   | [spec] | N | $XX |

### Warnings

- [High-cost or high-risk item]

### Optimization Opportunities

- [Potential saving and rationale]

Budgets, Tagging, Commitments, Optimization

Budget alerts

Use staged alerts at:

• 50%
• 80%
• 100%
• 120%

Required cost tags

Use these tags everywhere cost allocation matters:

• Project
• Environment
• Team
• CostCenter
• ManagedBy

Commitment strategy

• Start around 60-70% committed coverage for stable workloads.
• Keep the remaining 30-40% flexible with on-demand or spot capacity.
• Prefer flexible commitments before rigid family locks.

Optimization checklist

• Remove NAT gateways from low-risk non-prod environments when alternatives exist.
• Replace Multi-AZ databases in non-prod with single-AZ where safe.
• Release unattached EIPs, stale disks, empty load balancers, and old snapshots.
• Move old storage to cheaper classes.
• Scale Cloud Run to zero in non-prod when appropriate.
• Right-size after at least 2+ weeks of usage data.