Missing data vs zero

Zero is an observation: the provider or data source reported no usage in a period. Missing means no trustworthy observation exists for that period.

Usage Dashboard preserves this distinction:

• an observed zero can be plotted at zero;
• a missed poll, failed refresh, absent native bucket, or unavailable metric is a gap;
• failed refreshes retain last-known-good state but do not create zero history;
• providers without a quota are excluded from capacity totals rather than counted as 0% used.

Replacing gaps with zero would understate usage, distort burn rate and forecasts, and falsely improve coverage. A line crossing a chart gap should not be interpreted as confirmed inactivity.