LoopyLuci Google Cloud Well-Architected Framework — Cost Optimization
Provides a structured approach to optimize cloud costs while maximizing business value, following the FinOps lifecycle: Inform → Optimize → Operate.
Core Principles
- Align cloud spending with business value
- Foster a culture of cost awareness
- Optimize resource usage
- Optimize continuously
Cost Optimization Strategies
| Area | Strategy |
|---|---|
| Compute | Use CUDs for steady workloads, Spot VMs for fault-tolerant tasks |
| Storage | Lifecycle policies (move to Nearline/Coldline/Archive) |
| Networking | Keep traffic regional, use Standard Tier, Cloud CDN |
| Managed Services | Prefer Cloud Run, GKE Autopilot, serverless |
| Visibility | Cloud Billing reports, BigQuery billing export, Looker Studio |
Code Example: Storage Lifecycle Policy (gcloud)
# Set lifecycle policy on a bucket
gcloud storage buckets update gs://my-bucket --lifecycle-file=lifecycle.json
// lifecycle.json
{
"lifecycle": {
"rule": [
{
"action": {"type": "SetStorageClass", "storageClass": "NEARLINE"},
"condition": {"age": 30}
},
{
"action": {"type": "SetStorageClass", "storageClass": "COLDLINE"},
"condition": {"age": 90}
}
]
}
}
Common Pitfalls
- Retrieval fees: Nearline/Coldline/Archive have retrieval costs — size data access patterns first
- Over-provisioning: Right-size based on Recommender data, not guesses
- Missing labels: 100% of resources should be labeled for cost attribution
- No budgets: Always configure billing alerts and budgets
Verification Checklist
- All resources labeled with key metadata (env, team, app)
- BigQuery billing export enabled
- Budgets and alerts configured
- Rightsizing recommendations reviewed
- Storage lifecycle policies active
- CUD coverage optimized monthly