Skip to main content

Data Freshness

The Data Freshness page at /data-quality/freshness shows how recently each table was modified and flags anything past a configurable staleness threshold. Use it as a fast first-pass for "what's broken upstream?".

Workflow​

  1. Pick Catalog → Schema → optional Table
  2. Set Max stale hours (default 24)
  3. Click Check Freshness

The page calls GET /data-quality/freshness/{catalog} with optional schema/table/threshold filters.

Result cards​

Three counters across the top:

  • Fresh — last modified < threshold
  • Stale — last modified ≥ threshold
  • Unknown — no last_modified recorded (rare; usually means the table has never been written)

Datatable​

ColumnNotes
Table FQNClick to jump to Explorer
Last modifiedAbsolute timestamp
Hours since updateColour-coded: green if fresh, red if stale
Status iconGreen check / red alert / grey question
Status badgefresh / stale / unknown

Sort by hours-since-update to surface the most stale at the top.

Setting realistic thresholds​

Different tables have different cadences. Hard-coding one global threshold is noisy. Use SLA Dashboard to set per-table freshness rules — the Freshness page is then for ad-hoc spot-checks; the SLA page is for production monitoring.

API​

GET /data-quality/freshness/{catalog}?schema={schema}&table={table}&max_stale_hours=24

Returns one entry per table with last_modified, hours_since, status.