Pipeline

When each source last loaded, how many rows each layer holds, and the current result of every data-quality test.

Source freshness

What each layer holds

Test coverage

Currently failing

Nothing is failing, which on a green build is the expected state. This section is not broken: every one of the tests above holds, so every dbt_test__audit table is empty. If one were failing, this table would name it and point at the audit table holding the offending rows.

What the build costs

The three sections above describe the warehouse. This one describes the build — one row per node per dbt build, appended rather than replaced, so it is the one table on this page that accumulates.

The most recent build ran 562 nodes.

The data-quality layer is the build: tests and unit tests cost several times what building every model costs.