As DevOps engineers, we're great at building scalable infrastructure. But optimizing cost often falls through the cracks until finance sends an angry email.
Here's my practical cheat sheet for managing Azure costs — commands, tools, and strategies that work in production.
Quick Audit: Find Waste in 5 Minutes
Find idle VMs (Azure CLI):
List VMs with less than 5% average CPU (last 7 days)
az monitor metrics list \
--resource <vm-resource-id> \
--metric "Percentage CPU" \
--interval PT1H \
--aggregation Average \
--start-time $(date -d '7 days ago' +%Y-%m-%dT%H:%M:%SZ)
Find unattached disks:
List all unattached managed disks (you're paying for these!)
az disk list --query "[?managedBy==null].{Name:name, Size:diskSizeGb, SKU:sku.name, RG:resourceGroup}" -o table
Find unused public IPs:
az network public-ip list --query "[?ipConfiguration==null].{Name:name, RG:resourceGroup}" -o table
Top Cost-Saving Strategies (Ranked by Impact)
| Strategy | Savings | Effort | Time to ROI |
|---|---|---|---|
| Right-size VMs | 30-50% | Medium | 1-2 weeks |
| Reserved Instances | 30-72% | Low | Immediate |
| Azure Hybrid Benefit | Up to 85% | Low | Immediate |
| Auto-shutdown dev/test | 65% on those VMs | Low | 1 day |
| Spot VMs for batch | Up to 90% | Medium | 1-2 weeks |
| Storage tiering | 50-90% on storage | Low | 1 week |
Automate Cost Governance
Budget alerts (ARM template snippet):
{
"type": "Microsoft.Consumption/budgets",
"name": "monthly-team-budget",
"properties": {
"category": "Cost",
"amount": 500000,
"timeGrain": "Monthly",
"notifications": {
"Over75Percent": {
"enabled": true,
"operator": "GreaterThan",
"threshold": 75,
"contactEmails": ["team-lead@company.com"]
}
}
}
}
Auto-shutdown schedule (Azure CLI):
az vm auto-shutdown -g MyResourceGroup -n MyDevVM --time 1900 --timezone "India Standard Time"
Tools Worth Knowing
- Azure Cost Management + Billing — Native cost analysis and forecasting
- Azure Advisor — Right-sizing and RI recommendations
- Azure Resource Graph — Query resources at scale
- Power BI Cost Dashboard — Custom team-level visibility
- Terraform/Bicep — Infrastructure-as-Code with cost-aware defaults
When to Get External Help
If your Azure bill is consistently above ₹10L/month and you don't have a dedicated FinOps practice, it's worth engaging specialists. I've seen teams at Cloud 9 Infosystems reduce Azure costs by 40%+ for enterprise clients using a combination of the strategies above plus reserved instance planning and license optimization that's hard to do without deep Microsoft partnership expertise.
Key Metrics to Track Monthly
- Cloud Unit Economics: Cost per transaction / per user / per GB
- Coverage ratio: % of compute covered by RIs or Savings Plans
- Waste ratio: Idle resources / total resources
- Anomaly alerts: Unexpected spend spikes
If this was useful, follow me for more Azure infrastructure content. Next week: building a cost anomaly detection system with Azure Functions + Logic Apps.