Execution health · rule node-build-error
dbt models that failed to build
A Metacenta review checks this under the rule Models build without error. Everything below applies whether or not you ever commission one.
What this rule checks
From the results of your last dbt run or build, this rule flags every model, seed or snapshot that failed or errored.
Why it matters
Whatever the model feeds is either missing or serving its previous contents. Nothing in the data says which, so a dashboard can show last week's numbers as current.
How to fix it
Fix the models that no longer build. Fix the build error. Until it is fixed, treat everything downstream of the model as possibly stale.
When it is fine to leave
Rarely. A model that fails because it is being retired should be disabled or deleted rather than left failing, so the next failure stands out.
What we need to check it
run_results.json from a dbt run or dbt build, alongside manifest.json. When the run reached no verdict on any model, the rule declines rather than report a pass.