Billing & DBUs
The Billing page at /finops/billing tracks Databricks DBU consumption and Azure (or AWS / GCP) cloud costs across a rolling 30-day window. It's the answer to "how much is Clone-Xs and the lakehouse it touches actually costing us?".
Where the data comes from
- DBU consumption —
system.billing.usage(Databricks system table) - Cloud cost — Azure Cost Management API / AWS Cost Explorer / GCP Cloud Billing (configured per workspace)
- List vs. effective price — applies your workspace pricing tier and reservation discounts when available
Configure cloud billing in Settings → FinOps → Cloud Provider before this page renders cloud costs.
Summary cards
| Card | Meaning |
|---|---|
| Total DBUs | Sum of DBU consumption in window |
| List cost (DBX) | DBUs × list rate |
| Cloud cost | Azure / AWS / GCP cost |
| DBX % | Databricks share of total |
| Avg daily cost | Total / days |
Charts
- Daily DBU trend — 30-day line chart, hover for tooltip
- Cost comparison — bar chart of Databricks vs. cloud cost per day
Cost drivers
A bottom-of-page table breaks down cost by SKU:
- Job compute, all-purpose clusters
- SQL warehouse
- DLT pipelines
- Photon usage
- Serverless
Each row shows DBUs, cost, % of total. Sort by cost descending to find the biggest line items.
Tuning
- Window — change the lookback in
clxs.yamlunderfinops.billing_window_days - Currency — defaults to USD; override per region
API
GET /finops/billing?days=30
GET /finops/azure-costs?days=30 # or /aws-costs, /gcp-costs
Returns { daily_trend, total_dbus, total_cost, avg_daily_cost, currency } and the SKU breakdown.
Related
- Cost Breakdown — by-dimension view
- Compute Costs — drill into compute SKUs
- Storage Costs — drill into storage