Docs > Platform Observability > Getting Started with Cloud Cost
Getting Started with Cloud Cost
Overview
This guide introduces AppStatus Cloud Cost and covers:
- Spend broken down by service, cluster and workload
- Kubernetes cost derived from real resource consumption
- Cost shown next to the traffic it serves
- Threshold alerting on rate of change
- Read-only access to connected cloud accounts
What is Cloud Cost in AppStatus?
Cloud Cost brings spend into the same place as your performance data. Instead of a bill in one tool and latency in another, you see what a workload costs beside how much traffic it actually serves.
That pairing changes which questions you can answer. Total spend growing alongside traffic is not waste; spend growing while traffic is flat is a configuration change, and the least-used workload is more often the problem than the most expensive one.
Key capabilities:
- Cost attribution by service, cluster and workload
- Kubernetes cost derived from actual consumption
- Cost-per-request alongside absolute spend
- Change-rate thresholds rather than monthly-total alerts
- Read-only connections — nothing here needs write access
What you will see
What the cost breakdown looks like
Scroll the table sideways to see every column.
Row three is the point of this view: high spend, almost no traffic, and a jump nothing else explains.
Troubleshooting
Cost data is missing for an account
Confirm the account is connected in Cloud Integrations and that billing access was granted. Providers publish billing data on their own schedule, so the newest days can lag by a little.
Attribution is coarse — everything lands in one bucket
Attribution follows your tags. Workloads without consistent tags cannot be separated, so tag first and the breakdown sharpens immediately.
Spend jumped with no deploy
Look for a configuration change rather than a code one — a scaling policy, a retention setting, or a new region. Flat traffic with rising spend is nearly always configuration.
Operational Guidance
- Cost per request is more useful than total spend — growth is not the same as waste.
- The most expensive workload is not always the problem; the least-used one usually is.
- A spend jump with flat traffic means a config change, not demand.
Step-by-Step Setup
Cost data follows your cloud connections, so start there. The setup that gives this page its value is tagging — attribution can only be as sharp as the tags your workloads carry.
Before you start
- Cloud accounts connected in Cloud Integrations
- Billing read access granted on those accounts
- 1
Connect your cloud accounts
Connect each account in Cloud Integrations with read-only access. Nothing about cost collection requires permission to change anything.
WhereCloud Integrations → Add connection - 2
Confirm cost data arrives
Open Cloud Cost and confirm spend appears broken down by service. Providers publish billing data on their own schedule, so the most recent days can lag.
WhereCloud Cost → Overview - 3
Tag workloads consistently
Attribution follows your tags. Workloads without consistent tags collapse into one bucket, so tag first — the breakdown sharpens immediately afterwards.
WhereYour deployment manifestsTipThis is the difference between "cloud costs $4,000" and knowing which workload to fix.
- 4
Read cost against traffic
Open the breakdown and compare each workload against the requests it serves. The most expensive workload is not usually the problem; the least-used one often is.
WhereCloud Cost → Breakdown - 5
Alert on change, not total
Set a threshold on rate of change rather than the monthly total. Growth that tracks traffic is not waste; a jump with flat traffic is a configuration change worth catching the same week.
WhereCloud Cost → Thresholds
Configuration Options
Every option you can set, what each choice means, and what to pick. Use this as a reference while you fill in the form.
Attribution and alerting
| Field | Options | What it does | Recommended |
|---|---|---|---|
| Cloud account | Read-only connection | Source of billing data. | Read-only. Nothing here needs write access. |
| Attribution | Tags / labels | Maps cost onto workloads. | Tag consistently or the breakdown stays coarse. |
| Spend threshold | Absolute / rate | When a change is flagged. | Rate of change. A monthly total alert always arrives too late. |
Feature Reference
Every feature, where to find it in the app, and what it does. Use this when you know what you want to do but not where it lives.
| Feature | Where in app | Description |
|---|---|---|
| Cost breakdown | Cloud Cost → Breakdown | Spend by service, cluster and workload. |
| Kubernetes cost | Cloud Cost → Clusters | Derived from real resource consumption. |
| Cost per request | Cloud Cost → Breakdown | Spend read against the traffic it serves. |
| Change alerts | Alerts | Catch a jump the week it happens. |
Next Steps
Continue building your monitoring stack:
