Skip to content

Metrics explorer

The Metrics explorer shows how your cost evolves day by day, broken down by rating metric. It is the tool for questions like "when did storage start climbing?" or "which metric drove last week's increase?".

It reads the CloudKitty rating time series, so the figures are the billed rating data, not estimates.

Choosing a window

Pick a start and end date for the analysis. The chart then plots one point per day across that range. Shorter windows are good for investigating a recent spike; longer windows show seasonality and slow drift.

Reading the chart

  • The chart is interactive (built with Chart.js). Hover a day to see its values.
  • Each rating metric type is drawn as its own series, for example:
    • instance_up (running instances / compute)
    • volume.size (block storage)
    • image.size (image storage)
    • ip.floating (floating IPs)
  • You can toggle individual metrics on and off to isolate one contributor or compare two.
  • A counter shows the number of distinct metrics present in the window.

Multiple projects and regions

When your scope covers several projects or regions, the explorer aggregates their time series together (querying them in parallel) so you get one combined view.

How it differs from the dashboard hero

Dashboard hero Metrics explorer
Question How much this month, right now? How did cost move over a chosen range?
Shape A single MTD total plus a 30-day category chart A full daily series per metric, any date range
Granularity Compute / storage / network Individual rating metric types

Use the dashboard to know the number; use the metrics explorer to understand its shape and cause.

JSON API

For automation or your own reporting, the same data is available as structured JSON from the /api/metrics endpoint, returning the list of days, metric types, per-metric series, per-metric totals and a grand total.

Note

The API is scoped to your session and organisation, exactly like the UI. It returns the same figures you see on screen.

Where to go next